@chrono-meta/fh-gate 1.4.40 → 1.4.41

package/CLAUDE.md

@@ -246,6 +246,14 @@ post-compaction 2026-06-12). 2026-06-15+: headless `claude -p` draws from the ha
 subscription — prefer in-session Agent dispatch when the plan gate allows; take the headless fallback
 knowingly. Record sim results in the Axes 2–3 marker + sub-agent invocation log.
 
+**Measurement-integrity pre-flight** (when the sim/dispatch is a *cross-model measurement* — pinned to
+a specific tier, comparing model behaviors, or feeding a paper/published claim): consult
+`knowledge/shared/harness-core/measurement-integrity-checklist.md` first — pin the **display name** not
+a slug (silent fallback to a weaker model is a measured failure), take **reps ≥ 3** on any
+borderline/contested verdict (single draw = noise), and use a **discriminating** identity probe (a
+generic "OK" proves nothing about which model answered). The instrument must be verified before the
+measurement is trusted.
+
 **Floor-tier canary (optional pre-screen — token-free, *below* the Sonnet sim)**: a local model weaker
 than or comparable to Sonnet (e.g. `ollama run qwen3:8b` on the local host today; a cross-family local
 panel — qwen3.x:27b / gemma4:12b-qat / gpt-oss:20b / devstral — on a GPU host once its remote-exec path

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@chrono-meta/fh-gate",
-  "version": "1.4.40",
+  "version": "1.4.41",
   "description": "FH runtime adapters — run FH governance, skills, and agents via Claude or Codex with machine-parseable gates.",
   "license": "MIT",
   "keywords": [

package/plugins/fh-meta/skills/context-doctor/SKILL.md

@@ -81,6 +81,8 @@ When context is near the limit and you want to *preserve state* rather than rese
 □ Switching to a different track/project
 □ Error debugging has gone long and noise has accumulated
 □ Previous attempts are unrelated to current work
+□ Context has grown large (context rot — measurable quality degradation as input length grows,
+  well before the window fills; threshold varies by model/task. Chroma 2025 context-rot research)
 ```
 
 → Recommend immediate `/clear` and restart in fresh context.
@@ -196,6 +198,7 @@ Explicit invocation (`/context-doctor`) always runs regardless of suppress state
 - `/context-doctor`
 - "token waste", "session is slow", "reading the whole file", "claudeignore", "context cleanup"
 - "context diet", "memory audit", "CLAUDE.md is heavy", "MEMORY.md size"
+- "context engineering", "context rot", "context collapse"
 
 ### Natural Language Triggers (activates without internal vocabulary)
 
@@ -210,6 +213,7 @@ Also activates when an external user expresses without token/context terminology
 | "Seems like it keeps re-reading the same thing" | Repeated full-read detected | Step 2 (burst pattern) |
 | "Answers get weird as the session gets longer" | Accumulated context noise | Step 3 (/clear recommended) |
 | "Context is getting full", "context meter is high" | Approaching context limit | Step 3 — propose Wrap-then-Compact pattern |
+| "context engineering", "doing context engineering", "context rot setting in" | 2026 industry term for context discipline (Chroma 2025 / Anthropic) | Step 2 + Step 3 |
 
 ## Three-Doctor Loop Integration