Review findings from PR #333:
P1: _download_via_uri assumed output_video.uri is always files/<id>.
The API can return a full resource URI or a ready-made
.../files/<id>:download?alt=media download URL, which produced an
invalid poll path with a second :download appended. New
_file_id_from_uri() extracts the bare id from every documented shape;
regression tests cover the full-URL form plus a parametrized matrix of
URI shapes.
P2: docs/PROVIDERS.md still described the Google key as TTS + Imagen
only. The shared-key section now covers gemini_omni_video (model id,
~$0.10/sec pricing table, paid-tier-only, edit-turn billing note), and
the env snippet, provider-to-tool mapping, and capability coverage
tables include the new provider.
Add gemini_omni_video, a native Gemini API provider wrapping
gemini-omni-flash-preview via the Interactions API. Text-to-video,
image/reference-to-video with <FIRST_FRAME>/<IMAGE_REF_N> prompt tags,
and stateful edit_video turns via previous_interaction_id — the only
provider in the fleet that can refine a clip without regenerating it.
Reuses the existing GOOGLE_API_KEY / GEMINI_API_KEY, so one Google key
now unlocks images, TTS, and video.
- New Layer 3 skill .agents/skills/gemini-omni (prompting, edit-loop
rules, tag/timecode syntax, preview limits) sourced from official
Google docs; linked via agent_skills and the AGENT_GUIDE Layer 3 map
- ai-video-gen gains the Gemini API gateway row + editing pointer
- veo_video/sora_video fallback lists and video_selector agent_skills
reference the new provider; quality_score 0.85 with rationale
- Contract tests: registry discovery, selector routing, status from
env keys, uri + inline delivery, edit turns, typed image parts,
store=false editability, cost clamp
COGVIDEO_VARIANTS declares cogvideo-2b i2v=False (it is t2v-only), but
cogvideo_video advertised image_to_video + reference_image unconditionally
and the variant flag was never consulted. An image_to_video brief against
the 2B variant reached the diffusion pipeline and failed opaquely.
- Add is_operation_available(operation) that derives capability from the
variant table (the selector calls it without inputs, so it reports the
DEFAULT variant cogvideo-5b: t2v + i2v both True). This replaces an
implicit unconditional-True.
- Add an execute()-time guard that consults the CALLER's chosen variant
and fails fast with a clear error when it lacks the requested mode
(2B + image_to_video), instead of dropping into generate_local_video.
- Add _variant_for(inputs) helper shared by estimate_runtime / the guard.
Tests pin: the 2B premise (i2v=False), default-variant capability
reporting, fast-fail for 2B+i2v (generation never runs), and that 5B+i2v
still routes through to generate_local_video.
Refs: docs/REVIEW-image-to-video-voice.md §8 #4
Co-Authored-By: Claude <noreply@anthropic.com>
Three routing defects in video_selector, none previously covered by
tests (REVIEW §8 #3, #5, #7); plus the routing-test coverage itself (#10).
#3 Seedance dedup race
tool_by_provider keyed by provider STRING, so two tools legitimately
sharing provider="seedance" (seedance_video=fal, seedance_replicate)
collided — only the first-registered was ever selectable; the other
was invisible to the selector regardless of rank. Key selectable tools
by NAME instead; ranking picks the best of the shared-provider backends.
#5 preferred_provider had no score-gap gate
The selector returned the preferred provider on the first ranking match
no matter how far below the top it scored (the comment claimed "unless
drastically worse" but nothing enforced it). Add a configurable
preferred_provider_gap (default 0.15): honor the preference only when
its best ranked tool is within the gap of the overall top, else yield
to the top-ranked provider.
#7 fallback_tools appended image_selector unconditionally
The motion-required prohibition lived only in director skills, so a
direct caller could silently fall back to an image-only tool for an
image_to_video / reference_to_video brief. Add input-aware
fallback_tools_for(inputs) that drops image_selector for
motion-required operations; keep the static fallback_tools property
(with image_selector) for external consumers / contracts.
#10 routing coverage
First routing tests for video_selector: dedup reachability, the gap
gate (honored / ignored / configurable), motion-aware fallback, and
estimate_cost / estimate_runtime delegation. 13 tests, scoring patched
for determinism so they test routing logic, not the scorer.
Full tools + contracts suite green (638 passed, 6 skipped).
Refs: docs/REVIEW-image-to-video-voice.md §8 #3, #5, #7, #10
Co-Authored-By: Claude <noreply@anthropic.com>
Every premium video provider sets quality_score (seedance 0.95, runway /
higgsfield 0.9) so the scorer ranks them above stock/local options.
grok_video had none, so it was scored only on supports/stability flags
despite shipping native synchronized audio (lip-sync + dialogue + SFX
in a single generation pass) — likely under-ranked.
Set quality_score=0.9, on par with the other native-audio premium
providers. Add a regression pinning the field and its get_info() surface.
Refs: docs/REVIEW-image-to-video-voice.md §8 #6
Co-Authored-By: Claude <noreply@anthropic.com>
The bundled wan22-t2v-4step.json workflow loads the 14B FP8 diffusion
pair (wan2.2_t2v_high/low_noise_14B_fp8_scaled.safetensors), which
produce 16-channel latents, but its VAELoader referenced
wan2.2_vae.safetensors — the WAN 2.2 5B model's VAE, which expects
48-channel latents. Every T2V run therefore failed at VAEDecode with:
Expected tensor to have size 48 at dimension 1, but got size 16
Switch the workflow to wan_2.1_vae.safetensors, matching the 14B
models and the sibling wan22-i2v-4step.json, and update the T2V
required-models list in tools/video/comfyui_video.py to match so
preflight checks for the VAE that is actually used.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
video_compose.get_info() reported render_engines.ffmpeg as always
available, unlike the real availability checks used for remotion and
hyperframes. On a machine without ffmpeg on PATH, preflight would
falsely report ffmpeg as usable, letting render_runtime="ffmpeg" get
locked at proposal time only to fail at compose.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Pixabay rejects per_page outside 3-200 with HTTP 400. The stock_sources
adapter already clamped, but the PixabayVideo and PixabayImage tools
passed the value through raw, so callers using per_page < 3 got a 400.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The timeout handling only took effect on a direct _remotion_render() call. The
high-level execute(operation='render') path goes through _render(), which builds
a fresh remotion_inputs dict (edit_decisions, output_path, profile) and dropped
remotion_timeout_ms — so callers of the documented operation='render' path never
got the timeout passed to the Remotion CLI. Forward it there.
Adds a test exercising _render() (not just _remotion_render()) to cover the
high-level forwarding path.
Refs #217
The high-level Remotion render path hid the useful failure reason. run_command
runs with check=True + capture_output, so a non-zero exit raised
CalledProcessError whose str() is only 'returned non-zero exit status 1' — the
actual Remotion diagnostics in stderr were dropped. Catch CalledProcessError
and surface the stderr/stdout tail, and TimeoutExpired with an actionable hint.
Also add a creator-facing remotion_timeout_ms input, passed through as
Remotion's --timeout (headless-browser setup + delayRender). Slow browser
startup on restricted networks previously failed opaquely at the default 30s
with no way to raise it. The subprocess timeout is widened to match so
run_command does not kill Remotion before its own timeout fires.
Closes#217
The selector path previously hid the custom-workflow feature: video_selector
filtered tools on per-operation readiness (bundled WAN models) and both
selectors only chose ToolStatus.AVAILABLE providers, so comfyui_image/
comfyui_video — DEGRADED when bundled model metadata is missing — were
dropped even when the ComfyUI server was up and the caller supplied a full
workflow_json/workflow_path plus output_node.
- Add a custom-workflow readiness path to both selectors: when a custom
workflow is supplied, eligibility is based on server availability (status
!= UNAVAILABLE) for any provider advertising supports.custom_workflow,
not on bundled-model readiness. A custom workflow also restricts routing
to custom-workflow-capable providers, since the graph JSON is ComfyUI
specific.
- Expose workflow_json, workflow_path, output_node, workflow_name,
workflow_model, and workflow_model_stack in both selector schemas so
agents can discover the feature without bypassing the selectors.
- image_selector only forwards the workflow inputs to providers that
declare them.
- Add contract tests for the new eligibility path and schema exposure.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Client queries ComfyUI /object_info to discover installed models
(checkpoints, diffusion models, VAE, CLIP, LoRAs)
- Each tool declares its required models and checks them on execute()
- get_status() returns DEGRADED when server is up but models are missing
- Clear error messages tell the user exactly which models to download
- When COMFYUI_SERVER_URL is not set, error message tells the user to
configure it in .env instead of silently failing on localhost:8188
- 8 new tests covering URL config, error messages, and model requirements
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds three new BaseTool providers that delegate GPU work to a running
ComfyUI server via its REST API. This avoids the need to install
PyTorch/diffusers directly, which is critical on hardware where the
ecosystem hasn't caught up (e.g. NVIDIA Blackwell / DGX Spark, aarch64
+ CUDA 13.0).
New files:
- tools/_comfyui/client.py — shared REST client (submit/poll/download)
- tools/_comfyui/workflows/ — 4 bundled workflow templates
- tools/graphics/comfyui_image.py — FLUX 2 Dev NVFP4 text-to-image
- tools/video/comfyui_video.py — WAN 2.2 14B t2v + i2v (4-step LightX2V)
- tools/audio/comfyui_music.py — ACE-Step 3.5B music generation
- tests/contracts/test_comfyui_tools.py — 41 contract tests
- docs/comfyui-adapter-plan.md — design document
Zero changes to existing tools, selectors, registry, or pipelines.
Tools are auto-discovered and selectors pick them up via capability match.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Closes the remaining three atelier audit gaps.
F1 — atelier projects live under repo-root projects/<slug>/ (the
established OpenMontage convention), not under remotion-composer/.
_render_via_atelier now auto-stages source files into
remotion-composer/projects/<slug>/ via mtime-skip copy when the entry
isn't already under the composer tree. Originally tried a directory
junction, but webpack walks up from the entry's REAL location to
resolve node_modules — junctions get dereferenced and module resolution
fails. Copy keeps source-of-truth under projects/, is invisible to the
user's project view, and is cheap (typical project is a few small .tsx
files). Output stays wherever output_path says. effective_entry is
returned in tool data for transparency.
F3 — scripts/scaffold_atelier_project.py generates the engine
boilerplate (index.tsx / Root.tsx / Composition.tsx / art-direction.md /
props template / README) ONLY. No creative content emitted — the
Composition.tsx is a deliberately blank black placeholder with prominent
TODOs. This removes the friction that nudged agents back to templated
mode while making it impossible to ship the scaffold unauthored (review
correctly flags black frames / silent audio as 'revise'). Verified
end-to-end: scaffold → render → review fires the right warnings.
F5 — schemas extended to know about atelier:
- decision_log: 'composition_mode' added to category enum.
- proposal_packet.production_plan: composition_mode + art_direction.
- edit_decisions: composition_mode + bespoke{entry, composition_id,
art_direction, props_path, public_dir, scale, crf, concurrency}.
- final_review.checks: 'atelier' block (stock_reuse_detected,
offending_imports[], art_direction_declared, art_direction).
- final_review.recommended_action: 're_author' added.
- final_review.transcript_comparison: allow null word_accuracy /
transcript_matches_script (pre-existing latent bug surfaced here).
Both pass and fail shapes of the atelier render's final_review now
validate against the updated schema.
Closes the two biggest gaps from the atelier audit: hero bespoke renders
were skipping every automated quality gate, and the templated→atelier
inversion was advisory rather than enforced.
F6 — _render_via_atelier now runs the full _run_final_review (technical
probe + visual spotcheck + audio spotcheck + duration drift + transcript
comparison) plus two atelier-specific doctrine checks:
- stock_reuse_check: rglobs the bespoke project tree for imports from
the stock creative registry (src/components, src/Explainer, etc.).
Any hit fails the render with status="fail" and names the offending
file — the doctrine bypass is now mechanically caught.
- art_direction_check: requires edit_decisions.bespoke.art_direction.
Missing → warning surfaced in final_review.issues_found; forcing
function for step 1 of skills/meta/bespoke-composition.md.
Verified end-to-end on the compound-snowball proof render: positive case
status=pass, atelier checks clean; negative case (deliberate stock import
+ missing art_direction) returns status=fail with both findings named.
F7 — skills/meta/reviewer.md gets a "Composition Authoring Mode Review"
section that enforces the inversion at every stage:
- proposal: composition_mode must be a presented decision with both
options; hero briefs default to atelier; templated for hero requires
an explicit sanctioned reason.
- scene_plan/edit: any stock cut.type in an atelier piece is CRITICAL;
bespoke.{entry, composition_id, art_direction} required.
- compose: consumes final_review.checks.atelier from F6; mandates a
recorded distinctness review (the human taste call the tool can't
automate — its absence is itself CRITICAL).
- publish: blocks ship on any unresolved atelier finding.
Engine knowledge stays reusable; creative components don't.
Introduce a hand-authored, project-local Remotion render path that bypasses
the cut-schema and the stock scene-type registry, for hero/bespoke videos
that must look distinct from one another.
- video_compose: composition_mode="atelier" (or renderer_family="bespoke")
routes to _render_via_atelier, which renders a project-local entry under
remotion-composer/projects/<slug>/ with an optional per-project public_dir
(skips copying the bloated shared public/). No cut-schema, no stock registry.
- skills/meta/bespoke-composition.md: routing skill — art direction
(visual-style) -> motion principles (Disney 12) -> engine mechanics
(remotion-best-practices + stock components read only as a mechanics codex)
-> atelier render. Doctrine: reuse engine knowledge, never creative components.
- AGENT_GUIDE: "Composition Authoring Mode" (templated vs atelier); default
atelier for hero work; scene-type catalog reframed as a mechanics codex.
- animation-runtime-selector + INDEX: authoring-mode-first pointers.
- base_tool.run_command: decode subprocess output as UTF-8/replace (Windows
cp1252 crashed the reader thread on Remotion's Unicode progress output).
- .gitignore: remotion-composer/projects/ (throwaway bespoke compositions).
The compose target resolution was resolved from `profile` (and documented as
overridable via edit_decisions.metadata.compose_target) but the per-segment
scale/pad filter hardcoded 1920x1080. As a result, vertical profiles such as
`tiktok` / `youtube_shorts` / `instagram_reels` silently produced landscape
1920x1080 output instead of 1080x1920 — a silent dimension bug, no error raised.
- Use the resolved target width/height in the per-segment scale/pad filter.
- Implement the previously-stubbed `metadata.compose_target` extension point:
{"width", "height", "fit"} where fit="pad" (letterbox, default, unchanged
behavior) or fit="cover" (scale-to-fill + centre-crop, ideal for vertical).
- Default with no profile/target stays 1920x1080 (backward compatible).
Adds tests/tools/test_video_compose_vertical.py covering default landscape,
profile=tiktok vertical, and compose_target cover override.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
runway (estimate_cost/estimate_runtime/execute) defaulted to gen4_turbo and
higgsfield (execute) defaulted to kling_3.0, while both schemas advertise
seedance_2.0 as model.default. Omitting model under-quoted cost (runway 6x:
$0.25 vs $1.50) and silently generated a different model than advertised,
violating the Decision-Communication / cost-accuracy contract.
Root cause was a default duplicated across schema + 3 methods that drifted.
Collapse it to a single _DEFAULT_MODEL constant referenced everywhere.
Add tests/tools/test_provider_model_defaults.py to lock each tool's
estimate default to its schema default and guard the execute path.
On Windows, passing --props and the JSON path as two separate CLI
arguments causes Remotion to mis-parse the value due to platform quote
escaping, failing with "neither valid JSON nor a file path to a valid
JSON file". Switch to the --props=<path> equals form, which Remotion
recommends for file paths and which works consistently across
platforms.
Fixes#172
- Add get_torch_device() helper in _shared.py: cuda > mps > cpu
- Guard MPS detection for torch builds lacking torch.backends.mps
- Check both is_built() and is_available() for MPS
- Route load_diffusers_pipeline() to resolved device instead of hardcoded cuda
- Use float32 on CPU (float16 is emulated/unreliable), float16 on MPS, bfloat16 on CUDA
- Guard enable_model_cpu_offload() to CUDA-only; fall back to .to(device) on MPS
- Enable attention slicing for MPS memory safety
- Add inspect-based signature guard for device= arg on RealESRGANer/GFPGANer
- Update install_instructions on all LOCAL_GPU tools to mention MPS/Apple Silicon
Lines 219 and 227 imported upload_image_fal from _shared but called
the non-existent upload_image_to_fal, raising a NameError whenever
image_to_video or reference_to_video used local image paths.
Adds a transcript_comparison check to VideoCompose._run_final_review
that word-diffs the whisper/whisperx transcript against script.txt and
fails loudly when the TTS engine literally voiced punctuation tokens
(dot, dots, ellipsis, comma, dash, hyphen, period). Chirp3-HD did this
to ellipses in a production run and it slipped past review; now it
cannot.
Also corrects the "tiny background video" gotcha in
skills/core/hyperframes.md. After six renders of blaming HyperFrames
CSS, the real root cause was 640x360 Pexels sources combined with a
fit-and-pad pre-transform — HyperFrames was rendering the letterboxed
input faithfully. Gotcha now walks through the ffprobe diagnostic and
the scale-to-cover fix, and keeps the wrapper-div pattern for the
right reasons (aspect mismatch handling, not framework bug workaround).
Four regression tests cover the new check: punctuation-leak detection,
clean-audio false-positive guard, graceful skip when inputs missing,
and always-present transcript_comparison section.
Separates creative grammar (renderer_family) from technical engine
(render_runtime) so HyperFrames can stand alongside Remotion as a
first-class runtime instead of masquerading as a Remotion sub-case.
Locks runtime choice at proposal stage and enforces it end-to-end: the
schemas require it, video_compose routes by it, the reviewer fails
closed on silent swaps, and a parametrized contract test walks every
pipeline manifest to ensure each planning-stage skill explains the
conversation to the user. Adds hyperframes_compose (scaffold/lint/
validate/render/doctor/add_block), a playbook -> CSS style bridge, and
vendored HyperFrames Layer 3 skills from commit d291358, pinned via
PROVENANCE.md for future re-sync. Final_review now records
render_runtime_used and runtime_swap_detected so compose lies are
catchable after the fact.
Expands the Seedance 2.0 surface from fal.ai-only to every gateway the
OpenMontage toolset can reach, so users on Replicate / Runway /
Higgsfield automatically route cinematic briefs through Seedance 2.0
instead of the gateway's prior default.
- tools/video/seedance_replicate (new): REPLICATE_API_TOKEN adapter for
bytedance/seedance-2.0 and bytedance/seedance-2.0-fast. Same provider
family ("seedance"), quality_score=0.95, premium capability flags.
- tools/video/runway_video: add seedance_2.0 / seedance_2.0_fast to the
model enum as the new default, flag native_audio + multi_shot +
camera_direction + lip_sync in supports, point agent_skills at
seedance-2-0 Layer 3 skill. Note Runway Unlimited/Enterprise + non-US
gating in the model description.
- tools/video/higgsfield_video: same pattern. seedance_2.0 is now the
default model. Cost estimates calibrated to Higgsfield credit pricing
(~$0.50-$1.20 per 5s).
- tools/video/_shared: annotate HEYGEN_PROVIDERS seedance_pro/lite as
Seedance 1.x (2.0 on HeyGen flows through Video Agent / Avatar Shots
which need a separate wrapper — tracked as follow-up).
- .agents/skills/seedance-2-0: update provider matrix to show which
gateways are wrapped (fal.ai, Replicate, Runway, Higgsfield) vs. which
are still roadmap (HeyGen Video Agent, BytePlus, Freepik, aggregators).
The scoring engine already deduplicates by provider="seedance" so
whichever gateway the user has configured wins automatically. Agents
passing preferred_provider="seedance" to video_selector route to the
best available Seedance 2.0 path without naming a specific gateway.
Seedance 2.0 is now routed as the top-ranked video generator whenever a
premium gateway is available. Touches the tool layer, scoring engine,
cinematic pipeline, and both skill layers so discovery works from every
entry point.
- tools/video/seedance_video: BETA stability, quality_score=0.95, add
reference_to_video operation plus 9 img + 3 vid + 3 audio ceilings,
fix pre-existing upload_image_fal import
- tools/base_tool: surface optional quality_score / success_rate /
latency fields in get_info so the scorer can read them
- lib/scoring: fix reliability enum-vs-string bug that was pinning every
available tool to 0.0, switch to overlap coefficient so rich best_for
descriptions aren't penalized, add premium-cinematic feature bonus
- pipeline_defs/cinematic + cinematic asset-director: add pixabay_music
and freesound_music, restore pixabay-first music default
- cinematic compose-director: mandatory Remotion preflight at stage entry
- New Layer 3 .agents/skills/seedance-2-0/SKILL.md (8-part prompt
structure, multi-shot, lip-sync, reference-to-video, provider landscape)
- New Layer 2 skills/creative/prompting/seedance-prompting.md
- Update ai-video-gen, video-gen-prompting, AGENT_GUIDE, INDEX to flag
Seedance 2.0 as the preferred premium default and make the skill
discoverable from every routing path
6 API-based adapters: Pixabay Video, Coverr, NARA (U.S. National Archives),
Library of Congress, Pond5 Public Domain, Videvo. 5 scraper-based adapters
for sites without APIs: ESA, NOAA, Mixkit, Dareful, JAXA. All follow the
StockSource protocol and are auto-discovered — no tool code changes needed.
Updated scene-director, asset-director, and idea-director skills with
source routing guidance so agents know which provider to use for which
content type (e.g. nara for historical, noaa for ocean, esa for space).
Also includes: grok_video capability updates (native audio, 1-15s duration,
new aspect ratios), seedance_video tool (Seedance 2.0 via fal.ai),
ProductReveal Remotion composition registration.
Adds a lightweight alternative to the corpus_builder + clip_search pipeline
that skips CLIP embeddings, motion scores, and index files. Uses the same
StockSource adapter protocol so it works with all providers (Pexels,
Archive.org, NASA, Wikimedia, Unsplash). Asset-director skill now documents
both fast path (direct search) and standard path (corpus + CLIP retrieval).
Pipeline manifest updated to make corpus tools optional.
All of these changes came out of running the P1 and P2 documentary-montage
audits end-to-end and watching specific things break. Grouping them into
one commit because they share a theme: making the corpus builder and its
stock source adapters robust enough that a real brief can produce a
real corpus without special-casing.
corpus_builder.py + new clip_cache.py + test_clip_cache.py
Phase 1 of the shared-corpus architecture. Adds
~/.openmontage/clips_cache/ — a process-safe, LRU-evicted cache of
downloaded clip files keyed by clip_id. Before each candidate download,
corpus_builder asks the cache whether the bytes already exist on disk
from a previous run; on a hit it hard-links (or copies on cross-drive)
the blob into the caller's corpus dir and skips the network fetch. On
a miss it downloads as usual and ingests the fresh file. Motivation:
re-running the P1 audit after every tool fix was re-downloading gigs
of archive.org footage that had already been fetched in the prior run.
Cache faults never block the pipeline — they degrade gracefully to
normal downloads. The cache bubbles counters into the corpus_builder
return payload so the production report can show hit/miss/bytes-saved.
Default 20 GB cap, overridable via OPENMONTAGE_CACHE_MAX_GB.
Full test coverage: try_link, ingest, stats, LRU eviction, manifest
persistence, lock behavior. 23 tests, tmp_path-scoped.
stock_sources/archive_org.py
1. Three-strategy query cascade (phrase_prox_10 → distinctive_and →
distinctive_or). Motivation: natural-language documentary queries
against archive.org Solr were zeroing out — "1950s family watching
television" returned 0 hits because Solr's default multi-term AND
over-intersects. Walks strict to loose and returns the first
non-empty strategy. Stop words, source hints ("prelinger",
"archive", "footage"), and year tokens ("1950s") are excluded from
the distinctive-token picks since they don't correlate with
Prelinger title tokens.
2. 150 MB per-rendition size cap. archive.org routinely hosts
multi-hundred-megabyte h.264 masters and one 2 GB Prelinger item
poisons corpus build wall-time and disk. Within a format bucket
we now pick the largest rendition under the cap; if nothing fits
we fall through to the next format rather than dropping the item.
3. 180 s default max-duration ceiling when the caller hasn't set one
— archive.org is the only source that routinely hosts feature-
length material and a naive fan-out pulls them into corpora that
only ever want a few seconds per clip.
stock_sources/wikimedia.py
Parallel 3-strategy cascade (full → top2_or → single_best). Motivation:
Commons CirrusSearch also defaults to AND across multi-word queries;
our first P2 diagnostic pass returned 0 video results for 10/10
queries. Same stop-word / source-hint / year-token stripping as
archive_org so the two adapters stay symmetric.
test_stock_source_adapters.py
Rewrote the wikimedia query-builder tests against the new cascade
API. Added coverage for multi-word fallback + source-hint/year
stripping.
video_compose.py
Two small fixes for the Remotion renderer on Windows:
1. Resolve output_path to absolute before invoking the CLI so the
binary can write wherever the caller asked regardless of cwd.
2. Pass cwd=composer_dir to run_command so npx can find the local
Remotion binary under node_modules/.bin. Without this, Windows
npx returns "could not determine executable to run" because it
resolves .bin relative to the process cwd rather than the
script's parent.
New end-to-end pipeline for building thematic documentary montages from
a locally-indexed corpus of free stock footage (Pexels, Archive.org,
NASA). The agent builds a project-local corpus, CLIP-ranks candidates
per scene slot, edits with motion-aware arc logic, and composes via
ffmpeg. No paid APIs required for the full path.
Pipeline definition and director skills:
- pipeline_defs/documentary-montage.yaml: 5-stage manifest
(idea -> scene_plan -> assets -> edit -> compose)
- skills/pipelines/documentary-montage/: 6 director skills
(executive-producer + idea/scene/asset/edit/compose directors)
Corpus and retrieval infrastructure:
- tools/video/corpus_builder.py: multi-source stock fan-out with
resumable append-only corpus index
- tools/video/clip_search.py: CLIP ViT-B/32 retrieval —
rank_for_slot, find_similar_set, diversify, stats
- tools/video/stock_sources/: base + pexels + archive_org + nasa
adapters with a pluggable BaseStockSource contract
- lib/clip_embedder.py: CLIP wrapper
- lib/corpus.py: corpus schema, jsonl append/read, motion-score
caching
video_compose fix rolled in because any concat-based pipeline depends
on it:
- Replace ambiguous -to with -t duration (was double-trimming cuts)
- Force re-encode + normalize to 1920x1080 @ 30fps (was keyframe-
snapping with -c copy and breaking concat on mixed-source corpora)
- Add silent-audio anullsrc fallback for clips without an audio
stream
README: add Documentary Montage row to the pipeline table and bump
the pipeline count from 11 to 12.
- New: Higgsfield video provider with multi-model routing (Kling 3.0, Veo 3.1, Sora 2, WAN 2.5, Soul Cinema) and Soul ID character consistency
- Updated: Runway provider with gen4_aleph and gen3a_turbo models, proper pixel-ratio mapping, probe_output, watermark param, RUNWAYML_API_SECRET env var support
- Docs: Updated provider counts (12→13), tool counts (51→52), added Higgsfield setup/pricing sections across README, AGENT_GUIDE, ARCHITECTURE, and PROVIDERS
Remotion is now the default composition engine for ALL final renders
when available — video clips, images, mixed content. FFmpeg is only
used as fallback when Remotion is not installed or for standalone
operations (trim, transcode). Also fixes three FFmpeg fallback bugs:
profile + copy codec conflict, stream order mapping, and segment
seeking for audio-first containers.
