fix(ci): structural CI failures from cycles 45-47

Three required-job failures caught by `gh run list` per the cycle-48
tightened quality protocol (operator-asked, "how are you going to
enforce quality"):

1. **markdownlint** (cycle 46+) — MD004 ul-style violation in
   docs/releases/v2.39.0-claims/all-claims-evidence-map.md:52. A `+`
   bullet at column 3 was being interpreted as a nested list using
   plus-style; rewrote prose to use "with" instead.

2. **validate-three-gap-supergate** (cycle 42+) — compile-health child
   failed every CI run since the supergate landed because it requires
   .agents/defrag/latest.json which doesn't exist on greenfield CI
   runners (the nightly workflow generates it via `ao defrag` into a
   tmpdir + COMPILE_OUTPUT_DIR env). Sibling gates (council-coverage,
   flywheel-proof) already SKIP under similar conditions; compile-
   health now does too: if neither AGENTS_DIR/defrag/latest.json nor
   AGENTS_DIR/overnight/*/defrag/latest.json exists AND
   COMPILE_OUTPUT_DIR is unset, SKIP with "enforced by nightly
   workflow" message. The nightly workflow remains the canonical
   enforcer.

3. **registry-check** (cycle 45) — registry.json stale because cycle
   45 added 4 new skills/evolve/references/*.md files without
   regenerating. `bash scripts/generate-registry.sh` regenerated;
   counts now reflect current state (77 skills, 43 hooks, 4 stores,
   14 job types, 63 evals, 148 CLI commands).

Verification:
- markdownlint docs/releases/v2.39.0-claims/*.md — clean.
- bash scripts/check-three-gap-supergate.sh --gap=all — PASS locally.
- AGENTS_DIR=/tmp/no-such-dir bash scripts/check-three-gap-supergate.sh
  --gap=durable-learning — PASS with SKIP message for compile-health
  (simulates CI greenfield).
- bash scripts/generate-registry.sh --check — "OK: registry.json is up
  to date".

Out of scope:
- agentops-contract-canaries cycle-45 failure (aggregate=0.9549) — same
  fingerprint as the soc-l4yt closed-flake; doesn't reproduce on cycle
  46/47 CI runs or in local 5/5 bats. Not addressed here.
- security-toolchain-gate (advisory; exit 3 with 0 missing tools) —
  pre-existing, not introduced by this branch.

docs/releases/v2.39.0-claims/all-claims-evidence-map.md

@@ -49,7 +49,7 @@
 - **Claim:** mt-olympus city is a working multi-agent deployment of
   AgentOps.
 - **Surface:** `~/dev/mt-olympus/` (separate repo, gc supervisor
-  + bd dolt remote prefix `mo`/`hq`). Cross-fleet reference in
+  with bd dolt remote prefix `mo`/`hq`). Cross-fleet reference in
   `~/.claude/reference/bushido.md` "mt-olympus dolt" section.
 - **Gate:** Not gated by this repo's CI; cross-repo claim. Listed for
   completeness — fully verifying this requires inspecting mt-olympus

registry.json

@@ -1,13 +1,13 @@
 {
   "schema_version": 1,
-  "generated_at": "2026-05-10T19:28:50Z",
+  "generated_at": "2026-05-12T12:53:01Z",
   "summary": {
-    "skills": 75,
+    "skills": 77,
     "hooks": 43,
     "knowledge_stores": 4,
     "job_types": 14,
-    "eval_files": 61,
-    "cli_commands": 147
+    "eval_files": 63,
+    "cli_commands": 148
   },
   "surfaces": {
     "skills": [
@@ -185,7 +185,7 @@
         "path": "skills/evolve/",
         "has_skill_md": true,
         "has_references": true,
-        "reference_count": 12
+        "reference_count": 16
       },
       {
         "name": "grafana-platform-dashboard",
@@ -355,6 +355,14 @@
         "has_references": true,
         "reference_count": 5
       },
+      {
+        "name": "validate",
+        "tier": "judgment",
+        "path": "skills/validate/",
+        "has_skill_md": true,
+        "has_references": false,
+        "reference_count": 0
+      },
       {
         "name": "vibe",
         "tier": "judgment",
@@ -371,6 +379,14 @@
         "has_references": true,
         "reference_count": 4
       },
+      {
+        "name": "curate",
+        "tier": "knowledge",
+        "path": "skills/curate/",
+        "has_skill_md": true,
+        "has_references": false,
+        "reference_count": 0
+      },
       {
         "name": "domain",
         "tier": "knowledge",
@@ -1145,8 +1161,11 @@
       {
         "suite": "workbench",
         "path": "evals/workbench/",
-        "eval_count": 0,
-        "evals": []
+        "eval_count": 2,
+        "evals": [
+          "baseline-scorecard",
+          "scorecard-latest"
+        ]
       }
     ],
     "cli_commands": [
@@ -1242,6 +1261,10 @@
         "name": "corpus",
         "path": "cli/cmd/ao/corpus.go"
       },
+      {
+        "name": "corpus_snapshot",
+        "path": "cli/cmd/ao/corpus_snapshot.go"
+      },
       {
         "name": "curate",
         "path": "cli/cmd/ao/curate.go"

scripts/check-three-gap-supergate.sh

@@ -78,8 +78,20 @@ gap_durable_learning() {
     else
         echo "  SKIP  flywheel-proof (cli/bin/ao not built)"
     fi
-    run_gate "compile-health" \
-        "bash $REPO_ROOT/scripts/check-compile-health.sh" || fails=$((fails+1))
+    # compile-health requires .agents/defrag/latest.json (or an overnight
+    # fallback). On greenfield CI runners neither exists; the nightly
+    # workflow runs `ao defrag` and then check-compile-health.sh against
+    # a tmpdir output. The supergate is per-push and should SKIP rather
+    # than fail when the artifact is structurally unavailable.
+    local agents_dir="${AGENTS_DIR:-$REPO_ROOT/.agents}"
+    if [ -z "${COMPILE_OUTPUT_DIR:-}" ] \
+       && [ ! -f "$agents_dir/defrag/latest.json" ] \
+       && ! find "$agents_dir/overnight" -path '*/defrag/latest.json' -type f 2>/dev/null | grep -q .; then
+        echo "  SKIP  compile-health (no defrag artifact; enforced by nightly workflow)"
+    else
+        run_gate "compile-health" \
+            "bash $REPO_ROOT/scripts/check-compile-health.sh" || fails=$((fails+1))
+    fi
     return "$fails"
 }