docs: changelog and blog post for v0.10.0 (#4245)

* docs: changelog and blog post for v0.10.0

* fix(changelog): batch a release's commits and summarize them in parallel

The 0.10.0 range summarized to 37 entries, against 65 for the 145-commit release
before it: whole user-facing fixes were unlisted.

Measured on that range, holding model and prompt fixed, one call over the whole
release produces 120 entries against 152 for five batches — 40 commits get an
entry only when batched, against 8 only when not. The input was never the
constraint: a 245-commit release is ~8500 tokens of commit subjects in total, so
every arrangement fits the context window many times over. What batching bounds
is how many commits the model weighs at once before it starts summarizing them
away.

* Commits are batched by token count and the batches summarized concurrently,
  followed by a dedup pass that merges the same change described twice across a
  batch boundary. The dedup result is re-ordered by the release's own commit
  order and filtered to commit_ids that were in the input, so a dropped or
  invented id cannot reorder the changelog or point a reader at a commit that
  isn't there.

* max_completion_tokens=16000 was a *shared* budget on a reasoning model —
  reasoning bills against it, so a long commit list spends most of it before the
  first entry is emitted. Worth 8 entries on its own (37 -> 45). Raised to 100000.

* The grouping rule read as licence to merge unrelated fixes that touch the same
  area, and "focus on user-facing changes" as licence to drop anything whose
  commit title sounds internal. Both are now scoped; this was the largest single
  effect (45 -> 126).

Default model is now gpt-5.6-terra.

hindsight-dev/hindsight_dev/generate_changelog.py

@@ -12,9 +12,11 @@ import os
 import re
 import subprocess
 import sys
+from concurrent.futures import ThreadPoolExecutor
 from dataclasses import dataclass
 from pathlib import Path
 
+from hindsight_api.engine.token_encoding import count_tokens
 from openai import OpenAI
 from pydantic import BaseModel
 from rich.console import Console
@@ -624,16 +626,67 @@ def render_migrations_section(migrations: list[Migration]) -> list[str]:
     return lines
 
 
-def analyze_commits_with_llm(
-    client: OpenAI,
-    model: str,
+# A release's commits are summarized in token-sized batches rather than in one call.
+# One call over a whole release makes the model compress: 0.10.0's 220 commits came
+# back as 37 entries against 65 for the 145-commit release before it, with whole
+# user-facing fixes unlisted. A batch is small enough that every commit in it can be
+# considered on its own, and batches are independent, so they run concurrently.
+#
+# The budget is deliberately far below the model's input limit. Commit subjects are
+# short: a 245-commit release is ~8500 tokens in total, so any budget near the context
+# window is one batch and changes nothing. What is being bounded here is how much the
+# model is asked to hold at once before it starts summarizing away, not what fits.
+BATCH_TOKEN_BUDGET = 2000
+MAX_PARALLEL_BATCHES = 8
+# gpt-5.6-terra bills reasoning tokens against the response budget, so this has to
+# leave room for the reasoning that precedes the first entry, not just the entries.
+RESPONSE_TOKEN_BUDGET = 100000
+
+
+class PromptCommit(BaseModel):
+    """One commit as the summarization prompt shows it, and what batching counts tokens over."""
+
+    commit_id: str
+    message: str
+
+
+def _prompt_commit(commit: Commit) -> PromptCommit:
+    return PromptCommit(commit_id=commit.hash, message=commit.message)
+
+
+def batch_commits(commits: list[Commit], token_budget: int | None = None) -> list[list[Commit]]:
+    """Split commits into batches of at most ``token_budget`` tokens of commit text.
+
+    Batches follow the input order, so each one holds a contiguous run of the
+    release's history and the merged result stays in commit order. A single commit
+    over budget gets a batch to itself rather than being dropped or truncated.
+    """
+    # Read the module constant at call time rather than binding it as a default:
+    # a default is bound at import, so raising BATCH_TOKEN_BUDGET to compare batched
+    # against unbatched output silently measures two batched runs.
+    token_budget = BATCH_TOKEN_BUDGET if token_budget is None else token_budget
+    batches: list[list[Commit]] = []
+    current: list[Commit] = []
+    current_tokens = 0
+    for commit in commits:
+        tokens = count_tokens(json.dumps(_prompt_commit(commit).model_dump()))
+        if current and current_tokens + tokens > token_budget:
+            batches.append(current)
+            current, current_tokens = [], 0
+        current.append(commit)
+        current_tokens += tokens
+    if current:
+        batches.append(current)
+    return batches
+
+
+def _build_analysis_prompt(
     version: str,
     commits: list[Commit],
     file_diff: str,
-    integration: str | None = None,
-) -> list[ChangelogEntry]:
-    """Use LLM to analyze commits and return structured changelog entries."""
-    commits_json = json.dumps([{"commit_id": c.hash, "message": c.message} for c in commits], indent=2)
+    integration: str | None,
+) -> str:
+    commits_json = json.dumps([_prompt_commit(c).model_dump() for c in commits], indent=2)
 
     subject = f"the {integration} integration for Hindsight" if integration else f"release {version} of Hindsight"
 
@@ -647,7 +700,7 @@ def analyze_commits_with_llm(
         "its own package) — integrations are versioned and changelogged separately"
     )
 
-    prompt = f"""Analyze the following git commits for {subject} (an AI memory system).
+    return f"""Analyze the following git commits for {subject} (an AI memory system).
 
 For each meaningful change, create a changelog entry with:
 - category: one of "feature", "improvement", "bugfix", "breaking", "other"
@@ -655,27 +708,116 @@ For each meaningful change, create a changelog entry with:
 - commit_id: the commit hash from the input
 
 Rules:
-- Group related commits into a single entry if they're part of the same change
+- Group related commits into a single entry ONLY when they are literally parts of one
+  change (a fix and its follow-up, the same feature landed across two commits). Do not
+  merge distinct fixes into one summary because they touch the same area
+- Be complete rather than selective: every commit that changes what a user, operator or
+  API caller can observe gets its own entry. A bug fix in retain, recall, reflect,
+  consolidation, embeddings, the CLI, the control plane, a provider integration, the
+  Docker images or the Helm chart is user-facing even when its title reads as internal
 - Skip trivial changes (typo fixes, formatting, internal refactoring)
 - Skip repository-only changes: README updates, CI/GitHub Actions, release scripts, changelog updates, version bumps{skip_integrations_rule}
 - Focus on user-facing changes that affect the product functionality
 - Use the exact commit_id from the input (pick the most relevant one if grouping)
 - If no meaningful changes remain after filtering, return an empty list
 
+This is one batch of the release's commits, so judge each commit on its own; do not
+assume a change is absent because its other half is not in this batch.
+
 Commits:
 {commits_json}
 
 Files changed summary:
 {file_diff[:4000]}"""
 
+
+def _parse_entries(client: OpenAI, model: str, prompt: str) -> list[ChangelogEntry]:
     response = client.beta.chat.completions.parse(
         model=model,
         messages=[{"role": "user", "content": prompt}],
         response_format=ChangelogResponse,
-        max_completion_tokens=16000,
+        max_completion_tokens=RESPONSE_TOKEN_BUDGET,
     )
+    parsed = response.choices[0].message.parsed
+    return parsed.entries if parsed else []
 
-    return response.choices[0].message.parsed.entries
+
+def deduplicate_entries(
+    client: OpenAI,
+    model: str,
+    entries: list[ChangelogEntry],
+    commit_order: list[str],
+) -> list[ChangelogEntry]:
+    """Merge entries that describe the same change across batch boundaries.
+
+    Batches are contiguous, but a feature and its follow-up fix can still straddle
+    two of them, and two batches can describe one rollout twice. A deterministic
+    pass drops repeated commit_ids first — that needs no model — and the LLM pass
+    only has to catch the same change described in two different ways.
+
+    The result is re-ordered by the release's own commit order and filtered to
+    commit_ids that were actually in the input, so a dropped or invented commit_id
+    cannot reorder the changelog or point a reader at a commit that isn't there.
+    """
+    seen: set[str] = set()
+    unique: list[ChangelogEntry] = []
+    for entry in entries:
+        if entry.commit_id in seen:
+            continue
+        seen.add(entry.commit_id)
+        unique.append(entry)
+
+    if len(unique) > 1:
+        prompt = f"""These changelog entries were produced independently from separate batches of one
+release's commits, so the same change may appear more than once, worded differently.
+
+Return the entries to keep:
+- Drop an entry only when another entry describes the same underlying change. When two
+  entries describe one change, keep the clearer summary and its commit_id
+- Keep everything else exactly as it is: same summary text, same category, same commit_id
+- Do not merge distinct changes, do not reword what you keep, and do not invent entries
+
+Entries:
+{json.dumps([e.model_dump() for e in unique], indent=2)}"""
+        deduped = _parse_entries(client, model, prompt)
+        by_id = {e.commit_id: e for e in unique}
+        # Keep the model's text but never its idea of which commits exist.
+        unique = [e for e in deduped if e.commit_id in by_id] or unique
+
+    position = {commit_id: i for i, commit_id in enumerate(commit_order)}
+    return sorted(unique, key=lambda e: position.get(e.commit_id, len(position)))
+
+
+def analyze_commits_with_llm(
+    client: OpenAI,
+    model: str,
+    version: str,
+    commits: list[Commit],
+    file_diff: str,
+    integration: str | None = None,
+) -> list[ChangelogEntry]:
+    """Use LLM to analyze commits and return structured changelog entries."""
+    batches = batch_commits(commits)
+
+    if len(batches) == 1:
+        return _parse_entries(client, model, _build_analysis_prompt(version, batches[0], file_diff, integration))
+
+    console.print(f"[blue]Summarizing {len(commits)} commits in {len(batches)} batches...[/blue]")
+    with ThreadPoolExecutor(max_workers=MAX_PARALLEL_BATCHES) as pool:
+        results = list(
+            pool.map(
+                lambda batch: _parse_entries(
+                    client, model, _build_analysis_prompt(version, batch, file_diff, integration)
+                ),
+                batches,
+            )
+        )
+
+    entries = [entry for batch_entries in results for entry in batch_entries]
+    console.print(f"[blue]{len(entries)} entries before deduplication[/blue]")
+    deduped = deduplicate_entries(client, model, entries, [c.hash for c in commits])
+    console.print(f"[blue]{len(deduped)} entries after deduplication[/blue]")
+    return deduped
 
 
 def build_changelog_markdown(
@@ -773,7 +915,7 @@ def write_changelog(path: Path, header: str, new_entry: str, existing_releases:
 
 def generate_changelog_entry(
     version: str,
-    llm_model: str = "gpt-5.2",
+    llm_model: str = "gpt-5.6-terra",
 ) -> None:
     """Generate changelog entry for a specific version."""
     api_key = os.environ.get("OPENAI_API_KEY")
@@ -881,7 +1023,7 @@ For full release details, see [GitHub Releases](https://github.com/vectorize-io/
 def generate_integration_changelog_entry(
     integration: str,
     version: str,
-    llm_model: str = "gpt-5.2",
+    llm_model: str = "gpt-5.6-terra",
 ) -> None:
     """Generate changelog entry for a specific integration version."""
     if integration not in VALID_INTEGRATIONS:
@@ -1004,8 +1146,8 @@ def main():
     )
     parser.add_argument(
         "--model",
-        default="gpt-5.2",
-        help="OpenAI model to use (default: gpt-5.2)",
+        default="gpt-5.6-terra",
+        help="OpenAI model to use (default: gpt-5.6-terra)",
     )
     parser.add_argument(
         "--integration",

hindsight-dev/tests/test_generate_changelog_batching.py

@@ -0,0 +1,135 @@
+"""Tests for token-batched, parallel commit summarization and its dedup pass.
+
+The LLM calls themselves are stubbed: what these pin is the batching arithmetic
+and the guarantees the dedup pass makes about the model's output (order, and that
+a commit_id the model invented or dropped can't corrupt the changelog).
+"""
+
+import json
+
+from hindsight_api.engine.token_encoding import count_tokens
+
+from hindsight_dev.generate_changelog import (
+    BATCH_TOKEN_BUDGET,
+    ChangelogEntry,
+    Commit,
+    analyze_commits_with_llm,
+    batch_commits,
+    deduplicate_entries,
+)
+
+
+def _max_commit_tokens(batch: list[Commit]) -> int:
+    """A batch may exceed the budget by its last commit, which is added before the check."""
+    return max(count_tokens(json.dumps({"commit_id": c.hash, "message": c.message})) for c in batch)
+
+
+def _commits(n: int, message: str = "fix(api): something user-facing happened here") -> list[Commit]:
+    return [Commit(hash=f"{i:09x}", message=f"{message} (#{i})") for i in range(n)]
+
+
+class _StubClient:
+    """Returns one entry per commit named in the prompt, and records the calls."""
+
+    def __init__(self, dedup_result: list[ChangelogEntry] | None = None):
+        self.prompts: list[str] = []
+        self._dedup_result = dedup_result
+        parse = self._parse
+        self.beta = type(
+            "_Beta",
+            (),
+            {"chat": type("_Chat", (), {"completions": type("_C", (), {"parse": staticmethod(parse)})()})()},
+        )()
+
+    def _parse(self, *, model, messages, response_format, max_completion_tokens):
+        prompt = messages[0]["content"]
+        self.prompts.append(prompt)
+        if prompt.startswith("These changelog entries were produced independently"):
+            entries = self._dedup_result if self._dedup_result is not None else []
+        else:
+            payload = json.loads(prompt.split("Commits:\n", 1)[1].split("\n\nFiles changed", 1)[0])
+            entries = [
+                ChangelogEntry(category="bugfix", summary=c["message"], commit_id=c["commit_id"]) for c in payload
+            ]
+        parsed = response_format(entries=entries)
+        return type("_R", (), {"choices": [type("_C", (), {"message": type("_M", (), {"parsed": parsed})()})()]})()
+
+
+def test_batches_stay_within_the_token_budget_and_preserve_order():
+    commits = _commits(200)
+    batches = batch_commits(commits)
+
+    assert len(batches) > 1
+    assert [c for batch in batches for c in batch] == commits
+    for batch in batches:
+        assert sum(count_tokens(json.dumps({"commit_id": c.hash, "message": c.message})) for c in batch) <= (
+            BATCH_TOKEN_BUDGET + _max_commit_tokens(batch)
+        )
+
+
+def test_the_budget_is_read_at_call_time(monkeypatch):
+    """A default bound at import would make raising the constant a silent no-op."""
+    commits = _commits(200)
+    assert len(batch_commits(commits)) > 1
+
+    monkeypatch.setattr("hindsight_dev.generate_changelog.BATCH_TOKEN_BUDGET", 10**9)
+    assert batch_commits(commits) == [commits]
+
+
+def test_a_single_commit_over_budget_gets_its_own_batch():
+    huge = Commit(hash="deadbeef1", message="x " * BATCH_TOKEN_BUDGET)
+    batches = batch_commits([huge, *_commits(2)])
+
+    assert batches[0] == [huge]
+    assert len(batches) > 1
+
+
+def test_a_small_release_is_one_call_with_no_dedup_pass():
+    client = _StubClient()
+    entries = analyze_commits_with_llm(client, "m", "0.1.0", _commits(3), file_diff="")
+
+    assert len(client.prompts) == 1
+    assert [e.commit_id for e in entries] == [c.hash for c in _commits(3)]
+
+
+def test_a_large_release_fans_out_and_deduplicates():
+    commits = _commits(200)
+    client = _StubClient(dedup_result=[])  # model drops everything -> fall back to the union
+    entries = analyze_commits_with_llm(client, "m", "0.1.0", commits, file_diff="")
+
+    analysis_prompts = [p for p in client.prompts if p.startswith("Analyze the following")]
+    assert len(analysis_prompts) == len(batch_commits(commits))
+    assert sum(p.startswith("These changelog entries") for p in client.prompts) == 1
+    # Every commit survives, in release order, despite the empty dedup response.
+    assert [e.commit_id for e in entries] == [c.hash for c in commits]
+
+
+def test_dedup_keeps_release_order_and_rejects_unknown_commit_ids():
+    commits = _commits(3)
+    entries = [ChangelogEntry(category="bugfix", summary=c.message, commit_id=c.hash) for c in reversed(commits)]
+    kept = [
+        ChangelogEntry(category="bugfix", summary="merged", commit_id=commits[2].hash),
+        ChangelogEntry(category="bugfix", summary="hallucinated", commit_id="ffffffff0"),
+        ChangelogEntry(category="bugfix", summary="kept", commit_id=commits[0].hash),
+    ]
+    client = _StubClient(dedup_result=kept)
+
+    result = deduplicate_entries(client, "m", entries, [c.hash for c in commits])
+
+    assert [e.commit_id for e in result] == [commits[0].hash, commits[2].hash]
+
+
+def test_dedup_drops_a_repeated_commit_id_before_asking_the_model():
+    commits = _commits(2)
+    duplicated = [
+        ChangelogEntry(category="bugfix", summary="first wording", commit_id=commits[0].hash),
+        ChangelogEntry(category="bugfix", summary="second wording", commit_id=commits[0].hash),
+        ChangelogEntry(category="bugfix", summary="other", commit_id=commits[1].hash),
+    ]
+    client = _StubClient(dedup_result=[])
+
+    result = deduplicate_entries(client, "m", duplicated, [c.hash for c in commits])
+
+    assert [e.summary for e in result] == ["first wording", "other"]
+    sent = json.loads(client.prompts[0].split("Entries:\n", 1)[1])
+    assert len(sent) == 2

hindsight-docs/blog/2026-09-14-version-0-10-0.md

@@ -0,0 +1,129 @@
+---
+title: "What's new in Hindsight 0.10.0"
+description: Images and files as first-class retain content, a 3.2x faster request path, a prompt preview for every operation, typo-tolerant tag matching, knowledge pages and mental models in document transfers, and the retirement of the bank profile and background endpoints
+authors: [nicoloboschi]
+date: 2026-09-14
+hide_table_of_contents: true
+tags: [release]
+---
+
+import ZoomableMedia from '@site/src/components/ZoomableMedia';
+
+Hindsight 0.10.0 lets you retain a screenshot or a PDF in the same call as the prose around it, makes the request path several times cheaper, and shows you the exact prompt an operation would send before you spend a token on it. It also retires two endpoints that have been deprecated for several releases.
+
+<!-- truncate -->
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-release.png" alt="Hindsight 0.10.0 highlights: images and files as first-class retain content, a 3.2x faster request path with concurrent embedding batches and faster token counting, a prompt preview and retain prompt tester, typo-tolerant tag matching and open-vocabulary labels, knowledge pages and mental models in document transfers, and the breaking changes" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+- [**Images and files are first-class retain content**](#images-and-files-are-first-class-retain-content): `content` takes an ordered list of text, image and file blocks, and facts cite the attachment they came from.
+- [**More throughput out of the same box**](#more-throughput-out-of-the-same-box): A 3.2x faster request path, concurrent embedding batches, and cheaper token counting.
+- [**See the prompt before you send it**](#see-the-prompt-before-you-send-it): A preview endpoint for retain, consolidation and reflect, plus a retain prompt tester in the control plane.
+- [**Recall: typo-tolerant tags, and labels you don't have to enumerate**](#recall-typo-tolerant-tags-and-labels-you-dont-have-to-enumerate): A `fuzzy` tag resolve mode, and an open-vocabulary `multi-text` entity label.
+- [**Take the synthesized layer with you**](#take-the-synthesized-layer-with-you): Knowledge pages and mental models travel in a document transfer.
+- [**Every bank on one page**](#every-bank-on-one-page): A banks overview with bulk delete is now the control plane's home page.
+- [**Breaking changes**](#breaking-changes): Removed bank profile and background endpoints, no `curl` in the Docker images, 503 instead of an unbounded queue under load, and one extension hook removed.
+
+## Images and files are first-class retain content
+
+`content` now accepts an ordered list of text, image and file blocks instead of only a string. The extractor reads each attachment in the position it occupies — a chart between two paragraphs is read as the chart between those two paragraphs — and every read surface hands back the attachments behind what it returns. Provenance is per fact, not per chunk: a fact stated in the prose doesn't claim the screenshot next to it as evidence.
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-retain-blocks.png" alt="The control plane's Add Document dialog in Blocks mode: a support thread's text, the error screenshot attached as its own block, and the reply that follows, in the order they appear in the source" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-document-attachment.png" alt="The retained document in the control plane, with the error screenshot rendered inline at the point in the support thread where it was attached" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+A plain string behaves exactly as before, byte for byte. Idempotency by content hash, `update_mode=append`, chunk-delta re-extraction and `reprocess_document` all keep working, because blocks are flattened at the API boundary into one canonical body and the bytes live in the existing file storage, content-addressed by SHA-256.
+
+**Recall carries them through.** A recalled fact comes back with the attachments behind it, in the API response and in the control plane, so the screenshot a fact was read from is one click away from the answer instead of something you go looking for in the source document.
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-recall-attachment.png" alt="The Recall Analyzer returning a fact with the request ID, bank ID and error code that only appear in the screenshot, with that screenshot shown under Attachments this fact came from" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+One migration ships with this (`attachments`, `document_attachments`); both tables are new and empty, so it's quick on any size of deployment.
+
+## More throughput out of the same box
+
+**A 3.2x faster request path.** Both HTTP middlewares were rewritten as pure ASGI. Starlette's `BaseHTTPMiddleware` spawns a child task per request and pipes the response through a pair of memory-object streams; on cheap routes that machinery cost more than the endpoint. On a 2-CPU container, `/health/live` went from 2476 rps at p99 108ms to 7917 rps at p99 17ms, at lower CPU. Recall is unchanged — it's CPU-bound in the endpoint, nowhere near that ceiling — but every cheap route, health check and poll got a lot cheaper. Nothing was dropped: the HTTP metrics and the unknown-parameter reporting both moved rather than went away.
+
+**Embedding batches go out concurrently.** Every remote embedding provider walked its batches in a plain loop, so a retain held exactly one embedding request open at a time however much text it had. Batching and a bounded fan-out now live in one place, results are concatenated in input order whatever the completion order, and the same TEI server sustains 903 texts/s at one in-flight request against 2080 at eight.
+
+**Cheaper token counting.** Counting went through building a full list of token ids only to take its length — once per candidate fact, candidate chunk, source fact and reranker document. Token counting now runs on [toktok](https://github.com/vectorize-io/toktok), the BPE tokenizer we built for this as a drop-in replacement for tiktoken: it returns a count without materialising the ids, which takes the four counting stages of one recall from 34.2 ms to 5.0 ms, and a 77k-token document from 36 ms and 2.8 MB of allocation to 3 ms and effectively none.
+
+Alongside those, recall stops building trace payloads when nothing asked for a trace, config resolution stops deep-copying the global config on every call, and the mental-model staleness check no longer walks the bank.
+
+## See the prompt before you send it
+
+A bank's missions and strategies only mean something once you can see the prompt they land in. `POST /banks/{id}/prompts/preview` returns the messages retain, consolidation or reflect _would_ send, in send order — no LLM call, no writes, runtime data replaced by a fixed placeholder. Messages come back as blocks, and the active ones concatenate to exactly the text that would be sent; a block for a setting you've switched off comes back marked inactive, so an unset mission is still visible in the place it would occupy.
+
+The control plane adds a **prompt tester for retain** on top of it: edit the input, see the prompt, and run the extraction to see what facts come out.
+
+<ZoomableMedia>
+<video src="/img/blog/version-0-10-0-extraction-tester.mp4" poster="/img/blog/version-0-10-0-extraction-tester.jpg" autoPlay loop muted playsInline width="100%" style={{borderRadius: "8px"}} aria-label="The Extraction Tester: switching extraction rules to verbose and adding an entity label updates the prompt blocks on the left, while a test extraction on sample text shows the resulting facts on the right" />
+</ZoomableMedia>
+
+## Recall: typo-tolerant tags, and labels you don't have to enumerate
+
+**Fuzzy tag matching.** Tag filtering is exact array containment, so a caller filtering on `typsecript` loses the memory tagged `typescript` before ranking ever runs — and no amount of better ranking fixes that, because the match itself has to tolerate the misspelling. A `tag_groups` leaf now takes an optional `resolve` field: leave it at `exact` and nothing changes, set it to `fuzzy` and that leaf's tags are matched against the bank's tags by trigram similarity instead of literally. That is the entire API change — no new config, no new response field.
+
+It resolves `typescropt`→`typescript`, `kubernets`→`kubernetes` and `user:alcie`→`user:alice`, and deliberately doesn't resolve `mango`→`mongo`. One limit is worth knowing up front: similarity is length-sensitive, so a short tag has few trigrams and a single edit destroys most of them — `kakfa`→`kafka` does not resolve. Fuzzy matching is effective on descriptive tags and close to inert on very short ones, and that same property is what keeps unrelated short words apart.
+
+**Open-vocabulary entity labels.** Entity labels could classify a fact against a fixed vocabulary, or capture one free string. Neither covers the case where you want _several_ values that can't be enumerated when the bank is configured — the names a thing is known by, including abbreviations and alternative spellings; the tickets a fact cites; product codes.
+
+The new `multi-text` type is a list of strings with no fixed vocabulary, so the extractor writes as many values as the content warrants. Each value becomes its own entity and, with `tag: true`, a tag — so a bank can derive a classification from its own content and then filter on it at recall, without the caller ever supplying the vocabulary. Both client wrappers can express this now; the TypeScript one previously had no way to configure entity labels at all.
+
+## Take the synthesized layer with you
+
+A document transfer moved documents and observations — the raw layer. Everything synthesized on top of it, the knowledge pages and mental models a bank had spent real LLM budget building, stayed behind and had to be rebuilt from scratch on the other side.
+
+Pass `include_knowledge_base` on a whole-bank export (or tick the box in the control plane) and the archive carries the knowledge pages and mental models too, alongside the documents. It's opt-in and off by default, and only available when exporting a whole bank rather than a document list, because a page synthesized from the whole bank means nothing next to an arbitrary subset of it. The manifest reports what came along, so an archive tells you whether it holds a knowledge base before you import it.
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-export-knowledge-base.png" alt="The control plane's Export documents dialog with Include Knowledge Base ticked, which adds mental models and knowledge pages to a whole-bank export" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+The import side preserves what the export captured: mental-model content survives the round trip intact, and observations keep their freshness rather than arriving as though they'd all been recorded at import time.
+
+## Every bank on one page
+
+The control plane used to open on a "Welcome to Hindsight" placeholder, so surveying a server that runs several banks meant switching into each one. It now opens on a list of every bank, with its name, memory count, last activity and creation date, paged like the other tables.
+
+<ZoomableMedia>
+<img src="/img/blog/version-0-10-0-banks-overview.png" alt="The control plane home page listing eight memory banks with their memory counts, last activity and creation dates, a checkbox on each row for bulk delete, a search box and a Create new bank button" width="100%" style={{borderRadius: "8px"}} />
+</ZoomableMedia>
+
+**Delete in bulk.** Select rows and delete them together instead of visiting each bank's danger zone. One failed delete doesn't stop the rest: the banks that couldn't be deleted stay listed with their error, still selected, ready to retry.
+
+The header logo now links back to this page, and the bank selector pins the bank you're in to the top of the list. A server with no banks at all still shows the welcome card, now with a button to create one.
+
+## Also in this release
+
+- **A registry for extensions shipped outside the server.** `hindsight-extensions/` is now where extensions distributed separately live, with the Supabase tenant extension as its first entry (`hindsight-ext-supabase-tenant`) — which takes a third-party IdP's JWT client out of every Hindsight install's dependency graph. **If you use it, this is an upgrade step:** the built-in path `hindsight_api.extensions.builtin.supabase_tenant` no longer exists, so an install still pointing at it fails at startup with `ModuleNotFoundError`. Add the extension to your image and set `HINDSIGHT_API_TENANT_EXTENSION=hindsight_ext_supabase_tenant:SupabaseTenantExtension`; every `HINDSIGHT_API_TENANT_*` setting and the schema naming are unchanged. New alongside it: `StaticKeysTenantExtension`, env-configured per-user API keys with per-schema isolation, for deployments that want multi-user isolation without an identity provider at all.
+- **Meta Model API as a first-class LLM provider.** `HINDSIGHT_API_LLM_PROVIDER=meta` routes extraction, reflection and consolidation to the Muse family. Note that Muse Spark always reasons — `reasoning_effort: "none"` is rejected, and reasoning tokens bill against the output budget, so give the per-operation token limits headroom.
+- **Reflect stops inventing numbers for gaps.** Asked about a period the bank holds no data for, reflect could extrapolate a specific figure from neighbouring periods and present it as reliably inferred. It now says the value isn't recorded, while still drawing qualitative conclusions from what is. The same release stops mental-model and knowledge-page delta refreshes from overwriting stored content with what the newest batch alone says — a running count replaced by the batch's count, or an earlier recorded event erased because the batch didn't mention it.
+- **Bulk ingest no longer starves other banks.** The worker rotates its slots across banks, so one bank ingesting a large corpus can't hold every slot while everyone else waits.
+- **Listing mental models returns metadata by default.** `GET .../mental-models` defaulted to every model's full synthesized content, which bloats a caller's context and pulls a bank's whole synthesized knowledge in one call. It now defaults to `detail=metadata`; pass `detail=content` or `detail=full` to get content back. The MCP `list_mental_models` tool is metadata-only — read a specific model with `get_mental_model`.
+- A **Business Executive** bank template, **Prometheus ServiceMonitor** support and `extraContainers`/`extraInitContainers` in the Helm chart, and optional **CPU profiling** to the logs, controlled by environment variables.
+
+## Breaking changes
+
+**The bank profile and background endpoints are gone.** `GET`/`PUT /v1/default/banks/{bank_id}/profile` and `POST /v1/default/banks/{bank_id}/background` have been deprecated for several releases. They now answer `410 Gone` with the replacement call in the detail. The routes stay in the OpenAPI spec with unchanged signatures, so no generated SDK method disappears from under a caller — only the behaviour changes.
+
+Disposition traits and the reflect mission are bank configuration, and already were: read and write them through the bank config API. The display name these endpoints also returned is on the bank list. To make the config API a complete replacement, **reading** `GET .../config` is no longer gated on `HINDSIGHT_API_ENABLE_BANK_CONFIG_API` — that flag now gates only writes, because a bank must always be able to read its own resolved settings.
+
+The control plane and the CLI moved in the same change. One behaviour difference worth knowing if you scripted against it: `hindsight bank background` now _replaces_ the mission rather than LLM-merging into it, and warns that it does.
+
+**`curl` is no longer in the Docker images.** It had one in-image consumer — the standalone image's own readiness loop, now replaced — and it was the sole reason nine unfixable HIGH vulnerability findings shipped in every image. Hindsight's images define no `HEALTHCHECK`, so nothing breaks by default, but if your Compose `healthcheck` or a Kubernetes `exec` probe runs `curl` inside the Hindsight container, it will now fail with "command not found". Switch it to an HTTP probe (Kubernetes `httpGet` against `/health/live`) or run the check from outside the container.
+
+**Overloaded servers refuse instead of queueing.** Recall, reflect and retain now pass through admission control, on by default. When a worker is at capacity, a request waits for a slot up to a deadline — 30 s for recall, 5 s for reflect, 2 s for retain — and is then refused with `503` and a `Retry-After` header. Before, it queued without limit behind the concurrency cap, and under a burst the latency simply moved into that queue. The Python and TypeScript clients retry recall and reflect on `503`, honouring `Retry-After`. They don't retry retain, because a retried synchronous retain could write twice. So **if you run synchronous retain in bursts, handle `503`** — back off and retry, or switch to async retain, which returns as soon as the operation is queued and isn't subject to this limit. The limits are derived per worker from the CPU the process has; tune them with `HINDSIGHT_API_ADMISSION_{RECALL,REFLECT,RETAIN}_MAX_IN_FLIGHT` and `_MAX_WAIT_MS`, and set a lane's `MAX_IN_FLIGHT` to a negative number to turn it off.
+
+**The `count_chunks` hook is gone from the memory extension API.** Nothing called it. If a custom memory-store extension implements it, the method is now dead code and can be deleted.
+
+---
+
+0.10.0 carries a long list of fixes across retain, recall, embeddings, the CLI and the supported providers. See the [changelog](/changelog) for the full list.

hindsight-docs/src/components/ZoomableMedia.module.css

@@ -0,0 +1,52 @@
+.frame {
+  position: relative;
+  cursor: zoom-in;
+  margin-bottom: var(--ifm-leading);
+}
+
+.frame > img,
+.frame > video {
+  display: block;
+}
+
+.zoom {
+  position: absolute;
+  top: 0.6rem;
+  right: 0.6rem;
+  display: inline-flex;
+  align-items: center;
+  justify-content: center;
+  width: 2rem;
+  height: 2rem;
+  padding: 0;
+  border: none;
+  border-radius: 0.4rem;
+  background: rgba(0, 0, 0, 0.55);
+  color: #fff;
+  cursor: pointer;
+  opacity: 0;
+  transition: opacity 0.15s ease;
+}
+
+.frame:hover .zoom,
+.zoom:focus-visible {
+  opacity: 1;
+}
+
+/* Touch screens have no hover, so the button would never appear. */
+@media (hover: none) {
+  .zoom {
+    opacity: 1;
+  }
+}
+
+/* Letterbox the media in fullscreen instead of stretching or cropping it. */
+.frame > img:fullscreen,
+.frame > video:fullscreen {
+  width: 100%;
+  height: 100%;
+  object-fit: contain;
+  background: #000;
+  border-radius: 0 !important;
+  cursor: auto;
+}

hindsight-docs/src/components/ZoomableMedia.tsx

@@ -0,0 +1,64 @@
+import React, { Children, isValidElement, useRef, type ReactElement, type ReactNode } from "react";
+import styles from "./ZoomableMedia.module.css";
+
+interface ZoomableMediaProps {
+  /** A single <img> or <video>; its `src` is also the fallback when fullscreen isn't available. */
+  children: ReactElement<{ src?: string }>;
+}
+
+/**
+ * Wraps one image or video with a button that shows it full screen.
+ *
+ * Uses the browser's own Fullscreen API, so Esc and the browser's controls close
+ * it with no lightbox to maintain. iOS Safari only allows fullscreen on video, so
+ * where the element has no requestFullscreen the full-size file opens in a new
+ * tab instead — the next best way to read a screenshot at full resolution.
+ */
+export default function ZoomableMedia({ children }: ZoomableMediaProps): ReactNode {
+  const frame = useRef<HTMLDivElement>(null);
+  const child = Children.only(children);
+  const src = isValidElement(child) ? child.props.src : undefined;
+
+  const openFile = () => {
+    if (src) window.open(src, "_blank", "noopener");
+  };
+
+  const showFullScreen = () => {
+    const media = frame.current?.firstElementChild as HTMLElement | null;
+    if (media?.requestFullscreen) {
+      // Rejects when the page isn't allowed fullscreen (e.g. embedded in an iframe
+      // without allowfullscreen); fall back to the file rather than doing nothing.
+      media.requestFullscreen().catch(openFile);
+    } else {
+      openFile();
+    }
+  };
+
+  return (
+    <div ref={frame} className={styles.frame} onClick={showFullScreen}>
+      {child}
+      <button
+        type="button"
+        className={styles.zoom}
+        aria-label="View full screen"
+        title="View full screen"
+        onClick={(event) => {
+          // The frame's own click handler would otherwise fire it a second time.
+          event.stopPropagation();
+          showFullScreen();
+        }}
+      >
+        <svg viewBox="0 0 24 24" width="16" height="16" aria-hidden="true">
+          <path
+            d="M4 9V4h5M20 9V4h-5M4 15v5h5M20 15v5h-5"
+            fill="none"
+            stroke="currentColor"
+            strokeWidth="2"
+            strokeLinecap="round"
+            strokeLinejoin="round"
+          />
+        </svg>
+      </button>
+    </div>
+  );
+}

hindsight-docs/src/pages/changelog/index.md

@@ -9,6 +9,203 @@ import PageHero from '@site/src/components/PageHero';
 This page covers the Hindsight core (API, CLI, control plane). Each integration is released on its own
 cadence and has its own changelog — find it on the [integration's page](/integrations).
 
+## [0.10.0](https://github.com/vectorize-io/hindsight/releases/tag/v0.10.0)
+
+**Breaking Changes**
+
+- Mental-model list responses now return metadata by default rather than full model content.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/cdbartholomew.png?size=40" alt="@cdbartholomew" width="18" height="18" style={{borderRadius: "50%"}} />@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fb94ce034" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fb94ce034</a>
+- API runtime Docker images no longer include curl.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5f9bff905" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5f9bff905</a>
+- Removed the bank profile and background API endpoints.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/163fbb0ed" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>163fbb0ed</a>
+- Removed the `count_chunks` extension hook from the memory extension API.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d8b11907b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d8b11907b</a>
+
+**Features**
+
+- Configure separate timeout and retry limits for each model in a multi-LLM chain.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/cdbartholomew.png?size=40" alt="@cdbartholomew" width="18" height="18" style={{borderRadius: "50%"}} />@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/54fc68e70" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>54fc68e70</a>
+- API requests can wait briefly for capacity, with clients retrying transient admission limits.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6ac46e230" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6ac46e230</a>
+- Prometheus metrics now cover every API worker and report event-loop lag distributions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/2cd0561f1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>2cd0561f1</a>
+- Recall and read results now include the attachments supporting facts, including for store-owned banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/48b62ee08" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>48b62ee08</a>
+- The control plane keeps the currently selected bank at the top of the bank selector.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4af316e74" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4af316e74</a>
+- Add a control-plane button to pause ambient constellation motion.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/558017a84" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>558017a84</a>
+- Enable CPU profiling through environment configuration and emit profiles to application logs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/da0444a72" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>da0444a72</a>
+- Add a control-plane prompt tester for retain and prompt previews for every operation.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9e6d9e76c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9e6d9e76c</a>
+- Route retained items to an LLM chain member based on their metadata.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/864d92692" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>864d92692</a>
+- Added a control-plane bank overview with bulk deletion support.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7244098c0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7244098c0</a>
+- Support images and files as first-class inline content when retaining memories.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/280f09820" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>280f09820</a>
+- Added Meta Model API as a first-class LLM provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/30ce3b8d1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>30ce3b8d1</a>
+- Added fuzzy matching support for leaf tags within tag groups during recall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f747d96c3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f747d96c3</a>
+- Added an option to run database migrations in a separate subprocess.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a46783815" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a46783815</a>
+- Added multi-valued open-text entity labels.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/21c216004" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>21c216004</a>
+- Webhook deliveries now include timestamped SHA-256 signatures for verification.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d936d4931" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d936d4931</a>
+- Added separate credential directories for each Codex LLM provider configuration.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c42e6323e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c42e6323e</a>
+- Mental-model delta refreshes now receive the operation schema.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f22c36250" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f22c36250</a>
+- Added configurable dimensions and more reliable startup checks for LiteLLM embeddings.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/dd60a542e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>dd60a542e</a>
+- Added a CUDA-enabled standalone Docker deployment recipe.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a6a6c8e5a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a6a6c8e5a</a>
+- Exposed the complete mental-model trigger policy through MCP, including knowledge pages.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/56ba04204" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>56ba04204</a>
+- Added a CLI option to choose full or delta trigger mode when updating mental models.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/mdbenito" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/mdbenito.png?size=40" alt="@mdbenito" width="18" height="18" style={{borderRadius: "50%"}} />@mdbenito</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1cc2f9c10" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1cc2f9c10</a>
+- Allow configuring the PostgreSQL schema that provides pg_search extension functions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/0b6b5261a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>0b6b5261a</a>
+- Added Helm support for Prometheus Operator ServiceMonitors.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/aWN4Y25pa2EK" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/aWN4Y25pa2EK.png?size=40" alt="@aWN4Y25pa2EK" width="18" height="18" style={{borderRadius: "50%"}} />@aWN4Y25pa2EK</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a6f99c995" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a6f99c995</a>
+- Allow extra containers and init containers in Helm deployments for the API, worker, and control plane.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Adrastopoulos" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Adrastopoulos.png?size=40" alt="@Adrastopoulos" width="18" height="18" style={{borderRadius: "50%"}} />@Adrastopoulos</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eb3d1d587" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eb3d1d587</a>
+- Added Control Plane translations for knowledge-base and mental-model refinement workflows.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/35b98cb94" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>35b98cb94</a>
+- Document transfers can now export a knowledge base.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8d8dab346" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8d8dab346</a>
+- Added a Business Executive memory-bank template.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/benfrank241" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/benfrank241.png?size=40" alt="@benfrank241" width="18" height="18" style={{borderRadius: "50%"}} />@benfrank241</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b75fa6103" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b75fa6103</a>
+- Added a per-bank option to disable text search and use pure vector recall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cd8d4d399" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cd8d4d399</a>
+
+**Improvements**
+
+- Reduce CPU usage and latency on recall requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d9df6d2a4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d9df6d2a4</a>
+- Improve API throughput, especially for inexpensive routes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/2544a73c4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>2544a73c4</a>
+- Speed up retain processing by accelerating semantic-link calculation within a batch.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b04579481" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b04579481</a>
+- Reduce overhead when resolving Hindsight configuration.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3e1d47fdc" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3e1d47fdc</a>
+- Improve performance when resolving duplicate entities within a retain batch.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bc06bd051" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bc06bd051</a>
+- Reduce API overhead by avoiding unnecessary mental-model staleness checks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/78c10a382" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>78c10a382</a>
+- Reduced standalone Docker image dependencies by removing the NodeSource bootstrap.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/798aebd66" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>798aebd66</a>
+- Reduced control-plane-only Docker image dependencies by omitting unused curl.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d6ec007df" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d6ec007df</a>
+- Updated container images to use the active Node.js 24 LTS runtime.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c57331a94" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c57331a94</a>
+- Improved Docker image build performance by setting file ownership during copy operations.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/SharkyRawr" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/SharkyRawr.png?size=40" alt="@SharkyRawr" width="18" height="18" style={{borderRadius: "50%"}} />@SharkyRawr</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/01ed1e15f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>01ed1e15f</a>
+- Reduced runtime Docker image size by removing unused libxml2 packages.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/rschlek" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/rschlek.png?size=40" alt="@rschlek" width="18" height="18" style={{borderRadius: "50%"}} />@rschlek</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f9e91e21f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f9e91e21f</a>
+- Included operation attributes in exported tracing spans for better observability.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/koriyoshi2041.png?size=40" alt="@koriyoshi2041" width="18" height="18" style={{borderRadius: "50%"}} />@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/356aa6cd4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>356aa6cd4</a>
+- Improved remote embedding throughput by processing provider batches concurrently.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9697c6900" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9697c6900</a>
+- Reduced API startup time by avoiding unnecessary loading of LLM and document-processing dependencies.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e48a0c9a7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e48a0c9a7</a>
+- Accelerated token counting used during memory processing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bce43b8e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bce43b8e1</a>
+- Added recall-phase metrics for recalls answered directly from stored memories.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b746ad139" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b746ad139</a>
+- Improved tracing by naming API operations clearly and linking queued retains to their workers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ebcd88f97" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ebcd88f97</a>
+- Added detailed timing metrics for recall request phases.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b00cc4047" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b00cc4047</a>
+- Added timing metrics for operation validation hooks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/97ae1d08c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>97ae1d08c</a>
+- Improved configuration hot-path performance by retaining cached settings where freshness is not required.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c9591229e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c9591229e</a>
+- Hardened Docker runtime images by removing npm from runtime stages and expanding base-image security scanning.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/rschlek" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/rschlek.png?size=40" alt="@rschlek" width="18" height="18" style={{borderRadius: "50%"}} />@rschlek</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/466517ff9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>466517ff9</a>
+- Improved recall performance by fetching expanded observation graphs in a single database query.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/jervaise" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/jervaise.png?size=40" alt="@jervaise" width="18" height="18" style={{borderRadius: "50%"}} />@jervaise</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3bba2c01f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3bba2c01f</a>
+- Avoided unnecessary document retagging when a tags update makes no change.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/43f545e9b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>43f545e9b</a>
+- Improved retain throughput by reducing embedding-vector conversion overhead.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/0478c09c4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>0478c09c4</a>
+- Reduced unnecessary database queries when creating or accessing banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/059951ce1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>059951ce1</a>
+- Improved pg_search query handling by using the backend's native tokenization.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/600890c43" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>600890c43</a>
+- Meter mental-model content based on what is delivered rather than the endpoint used.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/cdbartholomew.png?size=40" alt="@cdbartholomew" width="18" height="18" style={{borderRadius: "50%"}} />@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/870d20d6f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>870d20d6f</a>
+- Improved CPU reranker performance by avoiding unnecessary full garbage collections.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6f9f935e6" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6f9f935e6</a>
+- Deployments can take ownership of database maintenance routines.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/cdbartholomew.png?size=40" alt="@cdbartholomew" width="18" height="18" style={{borderRadius: "50%"}} />@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7ba5fb23a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7ba5fb23a</a>
+- Reduced memory overhead while retaining large documents by accumulating document content across sub-batches.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3ad9d618e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3ad9d618e</a>
+- Improved tokenization performance and updated the default tokenizer encoding.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9fcb7ca7a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9fcb7ca7a</a>
+- Improved retain throughput by batching chunk embedding requests for each document.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ca2c983e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ca2c983e1</a>
+
+**Bug Fixes**
+
+- CPU profiling now captures worker processes as well as the supervisor when enabled.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cdac00842" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cdac00842</a>
+- Custom memory stores no longer maintain indexes that do not apply to them.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5a6083175" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5a6083175</a>
+- Reflect no longer states a specific value for a period the memories don't cover — it says the data isn't recorded instead of extrapolating a number from neighbouring periods.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/630c3a63e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>630c3a63e</a>
+- Mental-model and knowledge-page delta refreshes no longer overwrite stored content from what the newest batch alone says, such as replacing a running count or erasing an earlier recorded event.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/630c3a63e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>630c3a63e</a>
+- Correct public API issues uncovered by end-to-end system testing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/59e25d74f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>59e25d74f</a>
+- Generated API clients now receive downloaded binary content as bytes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/88456981f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>88456981f</a>
+- Entity mention counts are restored correctly when mentions are removed.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/64a98aa56" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>64a98aa56</a>
+- Clearing an occurrence start date also clears the legacy event date.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/203d95c7c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>203d95c7c</a>
+- Reflection turns use unique tool-call IDs for compatibility with strict LLM APIs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/gwthm-in" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/gwthm-in.png?size=40" alt="@gwthm-in" width="18" height="18" style={{borderRadius: "50%"}} />@gwthm-in</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5df6398fa" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5df6398fa</a>
+- Retain operations no longer risk chunk ID collisions between banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/179938a65" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>179938a65</a>
+- Reflection reports when structured-output extraction fails instead of silently obscuring the cause.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f5b3f76a8" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f5b3f76a8</a>
+- Expose the bank-template import request body in the API schema and generated clients.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1081a2ea4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1081a2ea4</a>
+- Store-owned retain operations no longer fail while recording their own log entries.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/134207d3c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>134207d3c</a>
+- List and graph API responses now return consistently typed records.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ef3ccdba3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ef3ccdba3</a>
+- Retain completion events are emitted only after the associated store transaction succeeds.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/kyletser" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/kyletser.png?size=40" alt="@kyletser" width="18" height="18" style={{borderRadius: "50%"}} />@kyletser</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/511c86e10" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>511c86e10</a>
+- Split append operations retain the full document body instead of only its trailing portion.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fda969777" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fda969777</a>
+- Recall traces now report the query timestamp supplied by the caller.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e366ff407" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e366ff407</a>
+- Vertex single-content embedding models now receive one text per embedding request.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b2a7257eb" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b2a7257eb</a>
+- Local ML ARM64 images now run on older ARMv8.0 CPUs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aeaf4b1bd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aeaf4b1bd</a>
+- Future-dated search items are safely handled in recency scoring.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/ebarkhordar.png?size=40" alt="@ebarkhordar" width="18" height="18" style={{borderRadius: "50%"}} />@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1ca8c4d60" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1ca8c4d60</a>
+- Sent Gemini embedding inputs as separate content items for correct provider handling.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a08ec8ddf" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a08ec8ddf</a>
+- Returned 404 for bank-scoped read requests targeting banks that do not exist.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/66992496f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>66992496f</a>
+- Limited embedding requests to each model's usable context window, including provider prefixes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a3c5b350d" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a3c5b350d</a>
+- Improved consolidation by explicitly identifying deletion targets and accounting for discarded batch responses.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/05c775c41" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>05c775c41</a>
+- Processed queued concurrent requests in arrival order to prevent unfair waiting.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/ebarkhordar.png?size=40" alt="@ebarkhordar" width="18" height="18" style={{borderRadius: "50%"}} />@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/310c9f694" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>310c9f694</a>
+- Applied configured token limits when reflection uses forced synthesis.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d39170010" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d39170010</a>
+- Prevented imports from creating observations that cite missing source units.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4a447c061" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4a447c061</a>
+- Loaded environment configuration before engine startup so configured concurrency limits take effect.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f19c424e0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f19c424e0</a>
+- Allow cancellation of operations that are still in progress.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b1de1b941" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b1de1b941</a>
+- Correctly recognized requests made through OpenAI-compatible LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/SharkyRawr" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/SharkyRawr.png?size=40" alt="@SharkyRawr" width="18" height="18" style={{borderRadius: "50%"}} />@SharkyRawr</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5092f3e11" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5092f3e11</a>
+- Sanitized all retained content at the memory engine boundary.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bd607ab8b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bd607ab8b</a>
+- Interpret "last weekend" requested during a weekend as the preceding weekend.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/NgoQuocViet2001" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/NgoQuocViet2001.png?size=40" alt="@NgoQuocViet2001" width="18" height="18" style={{borderRadius: "50%"}} />@NgoQuocViet2001</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cef202e6b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cef202e6b</a>
+- Prevented the CLI from aborting when its output is piped to a command that closes early.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/2anoubis" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/2anoubis.png?size=40" alt="@2anoubis" width="18" height="18" style={{borderRadius: "50%"}} />@2anoubis</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a72f9de15" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a72f9de15</a>
+- Added bounded retry handling for all remote reranker requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d69abe7b9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d69abe7b9</a>
+- Used one thread-local HTTP client per thread for TEI embeddings to avoid cross-thread client issues.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e7987ba69" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e7987ba69</a>
+- Applied the shared embedding retry policy to Cohere and ZeroEntropy providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c1f70087f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c1f70087f</a>
+- Applied the shared embedding retry policy to Gemini requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/36f4a061f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>36f4a061f</a>
+- Made split-synthesis reflection map calls deterministic by using zero temperature.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/98e565d3f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>98e565d3f</a>
+- Allowed Bedrock inference profile ARNs to be used through the LiteLLM embedding provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7a52b0a6b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7a52b0a6b</a>
+- Preserved per-memory entity-label tags when re-retaining content updates document tags.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/576cf40a2" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>576cf40a2</a>
+- Installed PostgreSQL extensions in the public schema for reliable database setup.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/901b5c696" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>901b5c696</a>
+- Forwarded all HINDSIGHT_* environment settings to the embedded daemon.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7eb0fcc61" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7eb0fcc61</a>
+- Ran tenant provisioning migrations within the database isolation boundary.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/216272814" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>216272814</a>
+- Showed the server response body in CLI output when API requests fail.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/737e5bf42" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>737e5bf42</a>
+- Paginated observation-scope and webhook listing endpoints.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d20893a83" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d20893a83</a>
+- Fixed fuzzy tag matching to compare tag values rather than namespaced tag names.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/00caa8eab" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>00caa8eab</a>
+- Fixed tokenizer cache leakage during memory processing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/53ed37488" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>53ed37488</a>
+- Aligned bank configuration behavior between client wrappers and the server.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e8518c392" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e8518c392</a>
+- Duplicate mental-model IDs now return a conflict response instead of failing unexpectedly.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/kubaodias" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/kubaodias.png?size=40" alt="@kubaodias" width="18" height="18" style={{borderRadius: "50%"}} />@kubaodias</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1a8a1f50e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1a8a1f50e</a>
+- Improved concurrent API reliability when initializing date parsing locales.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3d0a66602" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3d0a66602</a>
+- Prevented shutdown activity from being falsely reported as an event-loop stall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f0cf699cc" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f0cf699cc</a>
+- Improved reliability of embedded local models with misaligned model weights.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4e952b4e5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4e952b4e5</a>
+- Re-retaining a document with changed scope now invalidates outdated observations.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7be0f88ef" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7be0f88ef</a>
+- LLM timeout errors now identify the request phase in which they occurred.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3531e82f9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3531e82f9</a>
+- Reflection runs now fail when their retrieval tools fail, rather than continuing with incomplete results.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/921ae824a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>921ae824a</a>
+- Fixed prioritized recall ranking so boosts are applied consistently in rank space.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aee42254a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aee42254a</a>
+- Fixed reflection output to use the required document sections format.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9bfad10a4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9bfad10a4</a>
+- Prevented bulk ingestion in one bank from starving work in other banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b75e94191" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b75e94191</a>
+- Skipped unnecessary reflection loops when a mental-model scope has no content to process.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c507e70e3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c507e70e3</a>
+- Stopped retain extraction from deriving its narrator from a bank’s display name.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aabbe8b22" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aabbe8b22</a>
+- Fixed authorization handling for bank-template imports.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/17fbc4f14" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>17fbc4f14</a>
+- Improved compatibility of strict structured-output schemas that use JSON references.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/db87a7467" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>db87a7467</a>
+- Fixed consolidation batching to resolve observation scopes before processing tag sets.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ferrastas" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/ferrastas.png?size=40" alt="@ferrastas" width="18" height="18" style={{borderRadius: "50%"}} />@ferrastas</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eff3546ec" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eff3546ec</a>
+- Ensured banks are authorized and provisioned before legacy mutations are applied.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/24adcfce7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>24adcfce7</a>
+- Expanded native knowledge BM25 searches to return matches for any search term.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a373ffab6" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a373ffab6</a>
+- Memory edits now update their vector embedding together with their fields.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/31db3983a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>31db3983a</a>
+- Made Grafana dashboards compatible with older Grafana versions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/oldnicke" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/oldnicke.png?size=40" alt="@oldnicke" width="18" height="18" style={{borderRadius: "50%"}} />@oldnicke</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7083fd790" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7083fd790</a>
+- Bank-specific configuration reads now use fresh bank settings instead of stale process-cached values.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/689c9b694" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>689c9b694</a>
+- Reprocessing retained content now re-runs extraction instead of leaving it unchanged.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9a4be5e05" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9a4be5e05</a>
+- Bound concurrent local embedding work to prevent resource exhaustion during ONNX embedding requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7051c6e3b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7051c6e3b</a>
+- Applied reliable per-request deadlines to all LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7729396e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7729396e1</a>
+- Applied keyword minimum-score thresholds consistently across all text-search backends.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/78d46a718" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>78d46a718</a>
+- Updated standalone Docker images to remove fixed high-severity vulnerabilities.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/44f937631" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>44f937631</a>
+- Improved recall ranking for broad date expressions by scoring the full date period.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6a796f9f5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6a796f9f5</a>
+- Fixed malformed JSON handling for OpenAI-compatible LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/98e29817a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>98e29817a</a>
+- Correctly send consecutive Anthropic tool results together in a user message.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/zlguo1996" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/zlguo1996.png?size=40" alt="@zlguo1996" width="18" height="18" style={{borderRadius: "50%"}} />@zlguo1996</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/169c6ef58" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>169c6ef58</a>
+- Ensure complete mental-model documents are embedded whenever they are written.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fd2e0c61f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fd2e0c61f</a>
+- Allow reflection delta updates to add sections containing explicit IDs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6b8dfe74f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6b8dfe74f</a>
+- Accurately report visible output-token usage for OpenAI-compatible providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/72c9c8f9e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>72c9c8f9e</a>
+- Read embedding profiles and log files as UTF-8 regardless of the host locale.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/koriyoshi2041.png?size=40" alt="@koriyoshi2041" width="18" height="18" style={{borderRadius: "50%"}} />@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ff29e5ad7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ff29e5ad7</a>
+- Apply all memory writes from a consolidation response atomically.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ca38687cd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ca38687cd</a>
+- Report a clear truncation error when a non-streaming LLM completion exceeds its output limit.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/ebarkhordar.png?size=40" alt="@ebarkhordar" width="18" height="18" style={{borderRadius: "50%"}} />@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7ef98859e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7ef98859e</a>
+- Avoid unsupported occurred-date constraints on retain backends that do not support them.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nickanderson" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nickanderson.png?size=40" alt="@nickanderson" width="18" height="18" style={{borderRadius: "50%"}} />@nickanderson</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5c78b8132" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5c78b8132</a>
+- Preserve base64 padding in API keys read from CLI configuration files.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/2anoubis" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/2anoubis.png?size=40" alt="@2anoubis" width="18" height="18" style={{borderRadius: "50%"}} />@2anoubis</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ddf57744c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ddf57744c</a>
+- Enforce the dedicated bank-read permission when viewing mental-model history.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sword-Saint69" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sword-Saint69.png?size=40" alt="@Sword-Saint69" width="18" height="18" style={{borderRadius: "50%"}} />@Sword-Saint69</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ad2ffa499" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ad2ffa499</a>
+- Preserve complete retain items so API reprocessing faithfully replays the original request.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/shauneccles" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/shauneccles.png?size=40" alt="@shauneccles" width="18" height="18" style={{borderRadius: "50%"}} />@shauneccles</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7b6177b46" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7b6177b46</a>
+- Avoid repeatedly attempting database settings that PostgreSQL has rejected as invalid.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Andreymi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Andreymi.png?size=40" alt="@Andreymi" width="18" height="18" style={{borderRadius: "50%"}} />@Andreymi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eca881d0b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eca881d0b</a>
+- Honor configured LLM temperature settings during reflection.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/koriyoshi2041.png?size=40" alt="@koriyoshi2041" width="18" height="18" style={{borderRadius: "50%"}} />@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7874561b0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7874561b0</a>
+- Read embedding profile configuration and Control Center environment files as UTF-8.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/koriyoshi2041.png?size=40" alt="@koriyoshi2041" width="18" height="18" style={{borderRadius: "50%"}} />@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3e8226d1a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3e8226d1a</a>
+- Preserve intended HTTP validation status codes for webhooks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/r266-tech" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/r266-tech.png?size=40" alt="@r266-tech" width="18" height="18" style={{borderRadius: "50%"}} />@r266-tech</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8743b4594" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8743b4594</a>
+- Apply knowledge-base node updates atomically.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/29901d43b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>29901d43b</a>
+- Sanitize model-generated text consistently at all LLM boundaries.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/BrianMcBrayer" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/BrianMcBrayer.png?size=40" alt="@BrianMcBrayer" width="18" height="18" style={{borderRadius: "50%"}} />@BrianMcBrayer</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aab937e5f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aab937e5f</a>
+- Do not apply source-language rules when an output language is explicitly configured.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/feniix" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/feniix.png?size=40" alt="@feniix" width="18" height="18" style={{borderRadius: "50%"}} />@feniix</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ba0be9b30" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ba0be9b30</a>
+- Ensure SDK trigger updates modify only the settings explicitly supplied by callers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/55dafc27b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>55dafc27b</a>
+- Stop stalled consolidation tasks after a bounded idle period.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/Sanderhoff-alt.png?size=40" alt="@Sanderhoff-alt" width="18" height="18" style={{borderRadius: "50%"}} />@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/698620786" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>698620786</a>
+- Accept reflection delta responses provided as a top-level operation array.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/ebarkhordar.png?size=40" alt="@ebarkhordar" width="18" height="18" style={{borderRadius: "50%"}} />@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7fe6a5e79" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7fe6a5e79</a>
+- Honor configured LLM timeouts when using the native Ollama provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/MasterST1337" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/MasterST1337.png?size=40" alt="@MasterST1337" width="18" height="18" style={{borderRadius: "50%"}} />@MasterST1337</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6e0f044f2" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6e0f044f2</a>
+- Detect and recover worker tasks that are stuck without making stage progress.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/BrianHotopp" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/BrianHotopp.png?size=40" alt="@BrianHotopp" width="18" height="18" style={{borderRadius: "50%"}} />@BrianHotopp</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/532aa8698" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>532aa8698</a>
+- Add bounded retries, backoff, and timeouts for LiteLLM embedding providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/icculp" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/icculp.png?size=40" alt="@icculp" width="18" height="18" style={{borderRadius: "50%"}} />@icculp</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/906dbcdbd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>906dbcdbd</a>
+- Improve retry behavior for Text Embeddings Inference embedding requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/oldnicke" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/oldnicke.png?size=40" alt="@oldnicke" width="18" height="18" style={{borderRadius: "50%"}} />@oldnicke</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8916986f5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8916986f5</a>
+- Bound within-batch retain linking work to prevent large deltas from exhausting worker memory.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8833a7518" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8833a7518</a>
+- Improved OpenAI rate-limit handling by honoring the provider's reset timing after 429 responses.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/romanbsd" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/romanbsd.png?size=40" alt="@romanbsd" width="18" height="18" style={{borderRadius: "50%"}} />@romanbsd</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/20e66093d" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>20e66093d</a>
+- Fixed transferred mental models so their evidence references remain valid.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c143de2a9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c143de2a9</a>
+- Fixed the Knowledge view so page and mental-model data is restricted to the selected bank.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3a399343b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3a399343b</a>
+- Fixed full-text search ranking so each result is scored independently.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/mameikagou" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/mameikagou.png?size=40" alt="@mameikagou" width="18" height="18" style={{borderRadius: "50%"}} />@mameikagou</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/23c63e173" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>23c63e173</a>
+- Fixed retain processing for oversized items so sub-batches are rebuilt from the correct source span.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/nicoloboschi.png?size=40" alt="@nicoloboschi" width="18" height="18" style={{borderRadius: "50%"}} />@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d258f7b82" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d258f7b82</a>
+- Restore administrator identity sequence values after database restores.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/r266-tech" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/r266-tech.png?size=40" alt="@r266-tech" width="18" height="18" style={{borderRadius: "50%"}} />@r266-tech</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e82f670f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e82f670f</a>
+
+**Other**
+
+- Kept the llama.cpp prompt cache in RAM to avoid persisting potentially sensitive prompts.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}><img src="https://github.com/cdbartholomew.png?size=40" alt="@cdbartholomew" width="18" height="18" style={{borderRadius: "50%"}} />@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/676e2139e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>676e2139e</a>
+
+**Database Migrations**
+
+This release alters high-volume tables (`memory_units_bm25`). Migrations run on startup, so allow extra time on large deployments.
+
+- `e2f4a6c8b0d1` — Add attachments and document_attachments (inline retain attachments).<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><code>attachments</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-warning-darker)", marginLeft: "0.25em"}}>medium</span> <code>document_attachments</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-warning-darker)", marginLeft: "0.25em"}}>medium</span><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/pull/4077" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>#4077</a>
+- `f3a5b7c9d1e2` — Drop the unused memory_units_bm25 materialized view.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><code>memory_units_bm25</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-danger)", marginLeft: "0.25em"}}>high volume</span><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/pull/4357" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>#4357</a>
+
 ## [0.9.2](https://github.com/vectorize-io/hindsight/releases/tag/v0.9.2)
 
 **Features**

skills/hindsight-docs/references/changelog/index.md

@@ -9,6 +9,203 @@ import PageHero from '@site/src/components/PageHero';
 This page covers the Hindsight core (API, CLI, control plane). Each integration is released on its own
 cadence and has its own changelog — find it on the integration's page.
 
+## [0.10.0](https://github.com/vectorize-io/hindsight/releases/tag/v0.10.0)
+
+**Breaking Changes**
+
+- Mental-model list responses now return metadata by default rather than full model content.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fb94ce034" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fb94ce034</a>
+- API runtime Docker images no longer include curl.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5f9bff905" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5f9bff905</a>
+- Removed the bank profile and background API endpoints.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/163fbb0ed" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>163fbb0ed</a>
+- Removed the `count_chunks` extension hook from the memory extension API.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d8b11907b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d8b11907b</a>
+
+**Features**
+
+- Configure separate timeout and retry limits for each model in a multi-LLM chain.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/54fc68e70" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>54fc68e70</a>
+- API requests can wait briefly for capacity, with clients retrying transient admission limits.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6ac46e230" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6ac46e230</a>
+- Prometheus metrics now cover every API worker and report event-loop lag distributions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/2cd0561f1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>2cd0561f1</a>
+- Recall and read results now include the attachments supporting facts, including for store-owned banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/48b62ee08" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>48b62ee08</a>
+- The control plane keeps the currently selected bank at the top of the bank selector.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4af316e74" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4af316e74</a>
+- Add a control-plane button to pause ambient constellation motion.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/558017a84" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>558017a84</a>
+- Enable CPU profiling through environment configuration and emit profiles to application logs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/da0444a72" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>da0444a72</a>
+- Add a control-plane prompt tester for retain and prompt previews for every operation.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9e6d9e76c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9e6d9e76c</a>
+- Route retained items to an LLM chain member based on their metadata.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/864d92692" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>864d92692</a>
+- Added a control-plane bank overview with bulk deletion support.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7244098c0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7244098c0</a>
+- Support images and files as first-class inline content when retaining memories.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/280f09820" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>280f09820</a>
+- Added Meta Model API as a first-class LLM provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/30ce3b8d1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>30ce3b8d1</a>
+- Added fuzzy matching support for leaf tags within tag groups during recall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f747d96c3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f747d96c3</a>
+- Added an option to run database migrations in a separate subprocess.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a46783815" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a46783815</a>
+- Added multi-valued open-text entity labels.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/21c216004" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>21c216004</a>
+- Webhook deliveries now include timestamped SHA-256 signatures for verification.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d936d4931" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d936d4931</a>
+- Added separate credential directories for each Codex LLM provider configuration.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c42e6323e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c42e6323e</a>
+- Mental-model delta refreshes now receive the operation schema.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f22c36250" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f22c36250</a>
+- Added configurable dimensions and more reliable startup checks for LiteLLM embeddings.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/dd60a542e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>dd60a542e</a>
+- Added a CUDA-enabled standalone Docker deployment recipe.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a6a6c8e5a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a6a6c8e5a</a>
+- Exposed the complete mental-model trigger policy through MCP, including knowledge pages.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/56ba04204" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>56ba04204</a>
+- Added a CLI option to choose full or delta trigger mode when updating mental models.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/mdbenito" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@mdbenito</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1cc2f9c10" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1cc2f9c10</a>
+- Allow configuring the PostgreSQL schema that provides pg_search extension functions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/0b6b5261a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>0b6b5261a</a>
+- Added Helm support for Prometheus Operator ServiceMonitors.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/aWN4Y25pa2EK" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@aWN4Y25pa2EK</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a6f99c995" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a6f99c995</a>
+- Allow extra containers and init containers in Helm deployments for the API, worker, and control plane.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Adrastopoulos" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Adrastopoulos</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eb3d1d587" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eb3d1d587</a>
+- Added Control Plane translations for knowledge-base and mental-model refinement workflows.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/35b98cb94" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>35b98cb94</a>
+- Document transfers can now export a knowledge base.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8d8dab346" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8d8dab346</a>
+- Added a Business Executive memory-bank template.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/benfrank241" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@benfrank241</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b75fa6103" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b75fa6103</a>
+- Added a per-bank option to disable text search and use pure vector recall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cd8d4d399" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cd8d4d399</a>
+
+**Improvements**
+
+- Reduce CPU usage and latency on recall requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d9df6d2a4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d9df6d2a4</a>
+- Improve API throughput, especially for inexpensive routes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/2544a73c4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>2544a73c4</a>
+- Speed up retain processing by accelerating semantic-link calculation within a batch.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b04579481" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b04579481</a>
+- Reduce overhead when resolving Hindsight configuration.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3e1d47fdc" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3e1d47fdc</a>
+- Improve performance when resolving duplicate entities within a retain batch.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bc06bd051" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bc06bd051</a>
+- Reduce API overhead by avoiding unnecessary mental-model staleness checks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/78c10a382" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>78c10a382</a>
+- Reduced standalone Docker image dependencies by removing the NodeSource bootstrap.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/798aebd66" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>798aebd66</a>
+- Reduced control-plane-only Docker image dependencies by omitting unused curl.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d6ec007df" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d6ec007df</a>
+- Updated container images to use the active Node.js 24 LTS runtime.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c57331a94" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c57331a94</a>
+- Improved Docker image build performance by setting file ownership during copy operations.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/SharkyRawr" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@SharkyRawr</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/01ed1e15f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>01ed1e15f</a>
+- Reduced runtime Docker image size by removing unused libxml2 packages.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/rschlek" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@rschlek</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f9e91e21f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f9e91e21f</a>
+- Included operation attributes in exported tracing spans for better observability.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/356aa6cd4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>356aa6cd4</a>
+- Improved remote embedding throughput by processing provider batches concurrently.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9697c6900" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9697c6900</a>
+- Reduced API startup time by avoiding unnecessary loading of LLM and document-processing dependencies.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e48a0c9a7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e48a0c9a7</a>
+- Accelerated token counting used during memory processing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bce43b8e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bce43b8e1</a>
+- Added recall-phase metrics for recalls answered directly from stored memories.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b746ad139" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b746ad139</a>
+- Improved tracing by naming API operations clearly and linking queued retains to their workers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ebcd88f97" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ebcd88f97</a>
+- Added detailed timing metrics for recall request phases.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b00cc4047" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b00cc4047</a>
+- Added timing metrics for operation validation hooks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/97ae1d08c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>97ae1d08c</a>
+- Improved configuration hot-path performance by retaining cached settings where freshness is not required.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c9591229e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c9591229e</a>
+- Hardened Docker runtime images by removing npm from runtime stages and expanding base-image security scanning.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/rschlek" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@rschlek</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/466517ff9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>466517ff9</a>
+- Improved recall performance by fetching expanded observation graphs in a single database query.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/jervaise" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@jervaise</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3bba2c01f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3bba2c01f</a>
+- Avoided unnecessary document retagging when a tags update makes no change.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/43f545e9b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>43f545e9b</a>
+- Improved retain throughput by reducing embedding-vector conversion overhead.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/0478c09c4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>0478c09c4</a>
+- Reduced unnecessary database queries when creating or accessing banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/059951ce1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>059951ce1</a>
+- Improved pg_search query handling by using the backend's native tokenization.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/600890c43" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>600890c43</a>
+- Meter mental-model content based on what is delivered rather than the endpoint used.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/870d20d6f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>870d20d6f</a>
+- Improved CPU reranker performance by avoiding unnecessary full garbage collections.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6f9f935e6" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6f9f935e6</a>
+- Deployments can take ownership of database maintenance routines.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7ba5fb23a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7ba5fb23a</a>
+- Reduced memory overhead while retaining large documents by accumulating document content across sub-batches.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3ad9d618e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3ad9d618e</a>
+- Improved tokenization performance and updated the default tokenizer encoding.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9fcb7ca7a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9fcb7ca7a</a>
+- Improved retain throughput by batching chunk embedding requests for each document.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ca2c983e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ca2c983e1</a>
+
+**Bug Fixes**
+
+- CPU profiling now captures worker processes as well as the supervisor when enabled.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cdac00842" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cdac00842</a>
+- Custom memory stores no longer maintain indexes that do not apply to them.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5a6083175" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5a6083175</a>
+- Reflect no longer states a specific value for a period the memories don't cover — it says the data isn't recorded instead of extrapolating a number from neighbouring periods.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/630c3a63e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>630c3a63e</a>
+- Mental-model and knowledge-page delta refreshes no longer overwrite stored content from what the newest batch alone says, such as replacing a running count or erasing an earlier recorded event.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/630c3a63e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>630c3a63e</a>
+- Correct public API issues uncovered by end-to-end system testing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/59e25d74f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>59e25d74f</a>
+- Generated API clients now receive downloaded binary content as bytes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/88456981f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>88456981f</a>
+- Entity mention counts are restored correctly when mentions are removed.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/64a98aa56" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>64a98aa56</a>
+- Clearing an occurrence start date also clears the legacy event date.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/203d95c7c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>203d95c7c</a>
+- Reflection turns use unique tool-call IDs for compatibility with strict LLM APIs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/gwthm-in" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@gwthm-in</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5df6398fa" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5df6398fa</a>
+- Retain operations no longer risk chunk ID collisions between banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/179938a65" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>179938a65</a>
+- Reflection reports when structured-output extraction fails instead of silently obscuring the cause.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f5b3f76a8" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f5b3f76a8</a>
+- Expose the bank-template import request body in the API schema and generated clients.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1081a2ea4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1081a2ea4</a>
+- Store-owned retain operations no longer fail while recording their own log entries.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/134207d3c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>134207d3c</a>
+- List and graph API responses now return consistently typed records.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ef3ccdba3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ef3ccdba3</a>
+- Retain completion events are emitted only after the associated store transaction succeeds.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/kyletser" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@kyletser</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/511c86e10" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>511c86e10</a>
+- Split append operations retain the full document body instead of only its trailing portion.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fda969777" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fda969777</a>
+- Recall traces now report the query timestamp supplied by the caller.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e366ff407" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e366ff407</a>
+- Vertex single-content embedding models now receive one text per embedding request.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b2a7257eb" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b2a7257eb</a>
+- Local ML ARM64 images now run on older ARMv8.0 CPUs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aeaf4b1bd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aeaf4b1bd</a>
+- Future-dated search items are safely handled in recency scoring.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1ca8c4d60" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1ca8c4d60</a>
+- Sent Gemini embedding inputs as separate content items for correct provider handling.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a08ec8ddf" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a08ec8ddf</a>
+- Returned 404 for bank-scoped read requests targeting banks that do not exist.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/66992496f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>66992496f</a>
+- Limited embedding requests to each model's usable context window, including provider prefixes.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a3c5b350d" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a3c5b350d</a>
+- Improved consolidation by explicitly identifying deletion targets and accounting for discarded batch responses.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/05c775c41" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>05c775c41</a>
+- Processed queued concurrent requests in arrival order to prevent unfair waiting.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/310c9f694" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>310c9f694</a>
+- Applied configured token limits when reflection uses forced synthesis.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d39170010" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d39170010</a>
+- Prevented imports from creating observations that cite missing source units.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4a447c061" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4a447c061</a>
+- Loaded environment configuration before engine startup so configured concurrency limits take effect.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f19c424e0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f19c424e0</a>
+- Allow cancellation of operations that are still in progress.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b1de1b941" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b1de1b941</a>
+- Correctly recognized requests made through OpenAI-compatible LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/SharkyRawr" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@SharkyRawr</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5092f3e11" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5092f3e11</a>
+- Sanitized all retained content at the memory engine boundary.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/bd607ab8b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>bd607ab8b</a>
+- Interpret "last weekend" requested during a weekend as the preceding weekend.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/NgoQuocViet2001" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@NgoQuocViet2001</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/cef202e6b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>cef202e6b</a>
+- Prevented the CLI from aborting when its output is piped to a command that closes early.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/2anoubis" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@2anoubis</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a72f9de15" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a72f9de15</a>
+- Added bounded retry handling for all remote reranker requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d69abe7b9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d69abe7b9</a>
+- Used one thread-local HTTP client per thread for TEI embeddings to avoid cross-thread client issues.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e7987ba69" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e7987ba69</a>
+- Applied the shared embedding retry policy to Cohere and ZeroEntropy providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c1f70087f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c1f70087f</a>
+- Applied the shared embedding retry policy to Gemini requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/36f4a061f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>36f4a061f</a>
+- Made split-synthesis reflection map calls deterministic by using zero temperature.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/98e565d3f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>98e565d3f</a>
+- Allowed Bedrock inference profile ARNs to be used through the LiteLLM embedding provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7a52b0a6b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7a52b0a6b</a>
+- Preserved per-memory entity-label tags when re-retaining content updates document tags.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/576cf40a2" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>576cf40a2</a>
+- Installed PostgreSQL extensions in the public schema for reliable database setup.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/901b5c696" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>901b5c696</a>
+- Forwarded all HINDSIGHT_* environment settings to the embedded daemon.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7eb0fcc61" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7eb0fcc61</a>
+- Ran tenant provisioning migrations within the database isolation boundary.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/216272814" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>216272814</a>
+- Showed the server response body in CLI output when API requests fail.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/737e5bf42" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>737e5bf42</a>
+- Paginated observation-scope and webhook listing endpoints.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d20893a83" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d20893a83</a>
+- Fixed fuzzy tag matching to compare tag values rather than namespaced tag names.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/00caa8eab" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>00caa8eab</a>
+- Fixed tokenizer cache leakage during memory processing.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/53ed37488" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>53ed37488</a>
+- Aligned bank configuration behavior between client wrappers and the server.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e8518c392" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e8518c392</a>
+- Duplicate mental-model IDs now return a conflict response instead of failing unexpectedly.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/kubaodias" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@kubaodias</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/1a8a1f50e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>1a8a1f50e</a>
+- Improved concurrent API reliability when initializing date parsing locales.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3d0a66602" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3d0a66602</a>
+- Prevented shutdown activity from being falsely reported as an event-loop stall.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/f0cf699cc" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>f0cf699cc</a>
+- Improved reliability of embedded local models with misaligned model weights.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/4e952b4e5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>4e952b4e5</a>
+- Re-retaining a document with changed scope now invalidates outdated observations.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7be0f88ef" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7be0f88ef</a>
+- LLM timeout errors now identify the request phase in which they occurred.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3531e82f9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3531e82f9</a>
+- Reflection runs now fail when their retrieval tools fail, rather than continuing with incomplete results.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/921ae824a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>921ae824a</a>
+- Fixed prioritized recall ranking so boosts are applied consistently in rank space.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aee42254a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aee42254a</a>
+- Fixed reflection output to use the required document sections format.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9bfad10a4" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9bfad10a4</a>
+- Prevented bulk ingestion in one bank from starving work in other banks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/b75e94191" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>b75e94191</a>
+- Skipped unnecessary reflection loops when a mental-model scope has no content to process.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c507e70e3" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c507e70e3</a>
+- Stopped retain extraction from deriving its narrator from a bank’s display name.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aabbe8b22" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aabbe8b22</a>
+- Fixed authorization handling for bank-template imports.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/17fbc4f14" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>17fbc4f14</a>
+- Improved compatibility of strict structured-output schemas that use JSON references.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/db87a7467" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>db87a7467</a>
+- Fixed consolidation batching to resolve observation scopes before processing tag sets.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ferrastas" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@ferrastas</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eff3546ec" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eff3546ec</a>
+- Ensured banks are authorized and provisioned before legacy mutations are applied.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/24adcfce7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>24adcfce7</a>
+- Expanded native knowledge BM25 searches to return matches for any search term.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/a373ffab6" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>a373ffab6</a>
+- Memory edits now update their vector embedding together with their fields.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/31db3983a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>31db3983a</a>
+- Made Grafana dashboards compatible with older Grafana versions.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/oldnicke" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@oldnicke</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7083fd790" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7083fd790</a>
+- Bank-specific configuration reads now use fresh bank settings instead of stale process-cached values.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/689c9b694" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>689c9b694</a>
+- Reprocessing retained content now re-runs extraction instead of leaving it unchanged.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/9a4be5e05" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>9a4be5e05</a>
+- Bound concurrent local embedding work to prevent resource exhaustion during ONNX embedding requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7051c6e3b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7051c6e3b</a>
+- Applied reliable per-request deadlines to all LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7729396e1" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7729396e1</a>
+- Applied keyword minimum-score thresholds consistently across all text-search backends.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/78d46a718" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>78d46a718</a>
+- Updated standalone Docker images to remove fixed high-severity vulnerabilities.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/44f937631" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>44f937631</a>
+- Improved recall ranking for broad date expressions by scoring the full date period.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6a796f9f5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6a796f9f5</a>
+- Fixed malformed JSON handling for OpenAI-compatible LLM providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/98e29817a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>98e29817a</a>
+- Correctly send consecutive Anthropic tool results together in a user message.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/zlguo1996" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@zlguo1996</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/169c6ef58" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>169c6ef58</a>
+- Ensure complete mental-model documents are embedded whenever they are written.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/fd2e0c61f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>fd2e0c61f</a>
+- Allow reflection delta updates to add sections containing explicit IDs.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6b8dfe74f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6b8dfe74f</a>
+- Accurately report visible output-token usage for OpenAI-compatible providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/72c9c8f9e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>72c9c8f9e</a>
+- Read embedding profiles and log files as UTF-8 regardless of the host locale.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ff29e5ad7" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ff29e5ad7</a>
+- Apply all memory writes from a consolidation response atomically.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ca38687cd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ca38687cd</a>
+- Report a clear truncation error when a non-streaming LLM completion exceeds its output limit.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7ef98859e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7ef98859e</a>
+- Avoid unsupported occurred-date constraints on retain backends that do not support them.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nickanderson" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nickanderson</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/5c78b8132" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>5c78b8132</a>
+- Preserve base64 padding in API keys read from CLI configuration files.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/2anoubis" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@2anoubis</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ddf57744c" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ddf57744c</a>
+- Enforce the dedicated bank-read permission when viewing mental-model history.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sword-Saint69" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sword-Saint69</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ad2ffa499" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ad2ffa499</a>
+- Preserve complete retain items so API reprocessing faithfully replays the original request.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/shauneccles" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@shauneccles</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7b6177b46" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7b6177b46</a>
+- Avoid repeatedly attempting database settings that PostgreSQL has rejected as invalid.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Andreymi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Andreymi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/eca881d0b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>eca881d0b</a>
+- Honor configured LLM temperature settings during reflection.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7874561b0" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7874561b0</a>
+- Read embedding profile configuration and Control Center environment files as UTF-8.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/koriyoshi2041" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@koriyoshi2041</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3e8226d1a" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3e8226d1a</a>
+- Preserve intended HTTP validation status codes for webhooks.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/r266-tech" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@r266-tech</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8743b4594" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8743b4594</a>
+- Apply knowledge-base node updates atomically.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/29901d43b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>29901d43b</a>
+- Sanitize model-generated text consistently at all LLM boundaries.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/BrianMcBrayer" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@BrianMcBrayer</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/aab937e5f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>aab937e5f</a>
+- Do not apply source-language rules when an output language is explicitly configured.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/feniix" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@feniix</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/ba0be9b30" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>ba0be9b30</a>
+- Ensure SDK trigger updates modify only the settings explicitly supplied by callers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/55dafc27b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>55dafc27b</a>
+- Stop stalled consolidation tasks after a bounded idle period.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/Sanderhoff-alt" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@Sanderhoff-alt</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/698620786" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>698620786</a>
+- Accept reflection delta responses provided as a top-level operation array.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/ebarkhordar" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@ebarkhordar</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/7fe6a5e79" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>7fe6a5e79</a>
+- Honor configured LLM timeouts when using the native Ollama provider.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/MasterST1337" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@MasterST1337</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/6e0f044f2" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>6e0f044f2</a>
+- Detect and recover worker tasks that are stuck without making stage progress.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/BrianHotopp" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@BrianHotopp</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/532aa8698" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>532aa8698</a>
+- Add bounded retries, backoff, and timeouts for LiteLLM embedding providers.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/icculp" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@icculp</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/906dbcdbd" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>906dbcdbd</a>
+- Improve retry behavior for Text Embeddings Inference embedding requests.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/oldnicke" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@oldnicke</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8916986f5" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8916986f5</a>
+- Bound within-batch retain linking work to prevent large deltas from exhausting worker memory.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/8833a7518" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>8833a7518</a>
+- Improved OpenAI rate-limit handling by honoring the provider's reset timing after 429 responses.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/romanbsd" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@romanbsd</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/20e66093d" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>20e66093d</a>
+- Fixed transferred mental models so their evidence references remain valid.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/c143de2a9" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>c143de2a9</a>
+- Fixed the Knowledge view so page and mental-model data is restricted to the selected bank.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/3a399343b" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>3a399343b</a>
+- Fixed full-text search ranking so each result is scored independently.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/mameikagou" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@mameikagou</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/23c63e173" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>23c63e173</a>
+- Fixed retain processing for oversized items so sub-batches are rebuilt from the correct source span.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/nicoloboschi" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@nicoloboschi</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/d258f7b82" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>d258f7b82</a>
+- Restore administrator identity sequence values after database restores.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/r266-tech" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@r266-tech</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/e82f670f" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>e82f670f</a>
+
+**Other**
+
+- Kept the llama.cpp prompt cache in RAM to avoid persisting potentially sensitive prompts.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/cdbartholomew" target="_blank" rel="noopener noreferrer" style={{color: "var(--ifm-color-primary)", textDecoration: "none", display: "inline-flex", alignItems: "center", gap: "4px", verticalAlign: "middle"}}>@cdbartholomew</a><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/commit/676e2139e" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>676e2139e</a>
+
+**Database Migrations**
+
+This release alters high-volume tables (`memory_units_bm25`). Migrations run on startup, so allow extra time on large deployments.
+
+- `e2f4a6c8b0d1` — Add attachments and document_attachments (inline retain attachments).<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><code>attachments</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-warning-darker)", marginLeft: "0.25em"}}>medium</span> <code>document_attachments</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-warning-darker)", marginLeft: "0.25em"}}>medium</span><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/pull/4077" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>#4077</a>
+- `f3a5b7c9d1e2` — Drop the unused memory_units_bm25 materialized view.<span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><code>memory_units_bm25</code><span style={{fontSize: "0.75em", color: "var(--ifm-color-danger)", marginLeft: "0.25em"}}>high volume</span><span style={{color: "var(--ifm-color-emphasis-500)", margin: "0 0.3em"}}>·</span><a href="https://github.com/vectorize-io/hindsight/pull/4357" target="_blank" rel="noopener noreferrer" style={{fontFamily: "var(--ifm-font-family-monospace, monospace)", fontSize: "0.85em", color: "var(--ifm-color-emphasis-600)"}}>#4357</a>
+
 ## [0.9.2](https://github.com/vectorize-io/hindsight/releases/tag/v0.9.2)
 
 **Features**