@chrono-meta/fh-gate 3.1.0 → 3.1.1

package/.claude-plugin/marketplace.json

@@ -11,19 +11,19 @@
   "plugins": [
     {
       "name": "fh-meta",
-      "version": "3.1.0",
+      "version": "3.1.1",
       "description": "New in 2.2.0: BREAKING (gate): chamber step 6 now reads ACTUAL.md, not BUDGET.md — an in-flight chamber run whose actual cost sits in BUDGET.md blocks until the ACTUAL: line moves to tracks/_chamber/<slug>/ACTUAL.md (the runner prints the path). Why: BUDGET.md's pre-verdict hash IS the ordering witness, and step 6 hard-blocked until that same file changed, so every run that reached COMPLETE necessarily mutated a witnessed artifact and verify returned TAMPERED — the chamber's promotion condition was unsatisfiable by construction, not by strictness. Two roles (immutable witness / post-verdict calibration sink) had collided in one file; each was correct alone, so neither side's code showed the conflict. Also: ko-tech-writer Step 2/4-b scans are now calibration-backed (known-pair fixtures + reproducible command, shipped) — discrimination is proven, 'zero residue' is explicitly NOT; chamber lane suite 12 -> 33 including the runner x witness seam no test covered; chamber_run.sh now teaches the two-commit discipline (gate hashes and verdict hash must land in separate commits/PRs — it previously advised the opposite). New in 2.1.0: BREAKING (gate): `crossfamily: declined` in an Axes 2-3 marker now requires grounds naming a record path that RESOLVES on disk — bare `declined`, and `declined` justified by author judgment, are blocked at commit. Remedy: cite where the operator decision lives (e.g. `.. — operator declined sidecars, per knowledge/shared/rules/operational_adaptation.md`), or use `DEGRADED_PANEL_UNUSED` if a panel was reachable and you chose not to recruit it — which is what author judgment actually is. `declined` was the only enum value with no grounds requirement; a cross-family review then broke the first (vocabulary-grep) fix three ways — self-validating on the value's own token, vacuous keyword passes, and over-blocking real declinations in natural prose — so the check asserts a resolvable record instead of words. Also: standpoint axis gains `tier1b` (a STATIC read of a target repo, executed nothing) plus a decide-in-order procedure, after blind floor-tier sims graded pure cold-reads as `tier2` three rounds running; steel-quench Wave 1's sixth angle (gate-locality) gains the output-template row it never had, so a mandatory angle stops being structurally unreportable; verify-bidirectional gains category 5 (prescriptive doctrine statement); Sister Asset Protocol gains an active-adoption trigger; new resident doctrine — Mechanization Boundary, Local Execution First, Skeleton-not-Muscle, Expedition track, and this package's versioning policy. Hub meta-operations toolkit — 35 skills + 7 agents. New in 2.0.1: harness-doctor cadence hook, portability lint wired into pre-commit, branch_claim.sh claim-count-vs-tree-count warning, louder confidentiality-scan fail-open notice, fh-gate.sh missing-package.json survival, identity ① reclassified 🟢 (cross-harness adapters + relay argument channel). New in 1.4.53: `fh-codex-doctor` (npm bin) — Codex adapter drift scanner; reads the documented M1/M2/M3 skill tier map + skill/agent source and reports codex-native/adapter-required/claude-native/unclassified per unit, wired into `npm test`/`prepublishOnly` (fail-closed on unclassified Claude-native primitives). New in 1.4.49: steel-quench gains Step 0.6 Verdict-Invariance Probe (groundedness axis — a load-bearing judged gate's verdict must track behavior, not rubric phrasing; measured flip-count over cross-family paraphrases; arXiv:2605.06161 Policy Invariance anchor); multi_model_sidecar_strategy §Vendor-native harness (a model is strongest in its own vendor CLI — Claude/CC, GPT/codex, Gemini/Antigravity; a universal router degrades all of them, so it stays an autocomplete/QA sidecar, never orchestration); predelete_check.sh fail-closed rewrite; memory-hygiene A-TMA anchor. New in 1.4.48: phantom-quench + steel-quench gain external frontier anchors (arXiv:2607.02052 package-hallucination; arXiv:2607.02057 prompt-coverage-adequacy); README model-flat claim reframed from a per-release point-curve to structural invariants (operation flattens across tiers; depth tier-order fixed within a generation). New in 1.4.47: onboarding step ① surfaces the Mode D companion-store session-start load in the auto-read salience anchor (previously only in the local binding + rules, so a greeting could skip the load). New in 1.4.46: context-doctor command-output axis (route to rtk/proxy for verbose CLI stdout, complementing .claudeignore; risk-gated to token-scarce envs). New in 1.4.41: context-doctor 2026 trigger vocab (context engineering/rot/collapse) + phantom-citation hardening; hub measurement-integrity-checklist (cross-model measurement pre-flight: display-name pin/reps≥3/discriminating probe). New in 1.4.40: install-wizard queryable-wiki scaffold (INDEX + session-start read + R/W/C ingest). New in 1.4.39: auto-decorrelation (cross-family verifier sidecar recruitment) + video-ingest (capability-routed video ingestion). New in 1.4.x: verify-axis check-class taxonomy (mandatory-pass/measured/judged), no-reinvention Tier-0 inventory, 7-class failure taxonomy, Destructive-Op Gate, Wave-T (Temper), tier-floor governance, Mode D Model Notice, FC consent lane, default-Sonnet guidance. New in 1.3.0: public-surface-audit, field-harvest Mode B auto-trigger, 4-axis gate scope ext. Validated cross-CLI: Claude Code, Codex, Gemini.",
       "source": "./plugins/fh-meta"
     },
     {
       "name": "fh-commons",
-      "version": "3.1.0",
+      "version": "3.1.1",
       "description": "Project-agnostic utility skills — 5 skills (convergence-loop · deliberation · mcp-circuit-breaker · token-budget-gate · ko-tech-writer) + 1 agent (quench-challenger). Domain-independent utilities transplantable into any project.",
       "source": "./plugins/fh-commons"
     },
     {
       "name": "fh-qp",
-      "version": "3.1.0",
+      "version": "3.1.1",
       "description": "QP (Quality Platform) — the generic edition of a field QA harness's Prepare→Automation→Regression loop as an FH plugin: 4 skills (qp router · qp-plan · qp-run · qp-regress) + qp_tools.sh (target-class · adapter-probe · mask · surface-reach · mtm-check · run-verbs, typed exit codes) + a zero-domain-constant profile slot + 29 known-pair lanes. Drives web targets through the session's Playwright MCP and desktop targets through computer-use MCP (mobile deferred); calls a registered qasp typed capability when one exists (strictest-wins) — none is registered today, so the MCP fallback is the first edition. Verdict contract: a MACHINE closure requires a recorded assertion; a failed first step is attributed BLOCKED, not FAIL; surface_reach counts every TC in the denominator. Born as chamber run #18 (EMIT, 2026-09-05).",
       "source": "./plugins/fh-qp"
     }

package/README.ja.md

@@ -58,7 +58,7 @@ brew tap chrono-meta/forge-harness && brew install forge-harness   # あるい
 **GitHub Actions では** — 同じゲートを1つのステップとして、判定は型のあるまま:
 
 ```yaml
-- uses: chrono-meta/forge-harness@v3.1.0
+- uses: chrono-meta/forge-harness@v3.1.1
   with:
     files: ${{ steps.changed.outputs.files }}
   env:

package/README.ko.md

@@ -45,7 +45,7 @@ brew tap chrono-meta/forge-harness && brew install forge-harness   # 또는 이
 **GitHub Actions 에서** — 같은 게이트를 스텝 하나로, 판정은 타입을 유지한 채:
 
 ```yaml
-- uses: chrono-meta/forge-harness@v3.1.0
+- uses: chrono-meta/forge-harness@v3.1.1
   with:
     files: ${{ steps.changed.outputs.files }}
   env:

package/README.md

@@ -47,7 +47,7 @@ brew tap chrono-meta/forge-harness && brew install forge-harness   # or this
 **In GitHub Actions** — the same gate as a step, with the verdict kept typed:
 
 ```yaml
-- uses: chrono-meta/forge-harness@v3.1.0
+- uses: chrono-meta/forge-harness@v3.1.1
   with:
     files: ${{ steps.changed.outputs.files }}
   env:

package/README.zh.md

@@ -57,7 +57,7 @@ brew tap chrono-meta/forge-harness && brew install forge-harness   # 或者用
 **在 GitHub Actions 里** —— 同一道门禁作为一个 step，判定依然是带类型的：
 
 ```yaml
-- uses: chrono-meta/forge-harness@v3.1.0
+- uses: chrono-meta/forge-harness@v3.1.1
   with:
     files: ${{ steps.changed.outputs.files }}
   env:

package/knowledge/shared/learnings/subagent_invocations_log.yaml

@@ -3517,3 +3517,13 @@
   evidence: "PR #264·265·266·268 전부 머지(main=01dafcd) · CI 6/6 × 4. cross-family 1차 A급 4건(3 수리 · 1 **반증** — needs_clarification 은 disposition 게이트가 이미 막는다) · 2차 S1+A4(전부 수리). 거버너 독립 되돌림 재검 4건 전부 적색 확인. 전체 스위트 1 failed/3996 passed(시간의존 픽스처 1, 기존)"
   tokens_subagent: 986241
   notes: "🟥 계기 확정 하나가 부수 수확: 「기존 실패 13건」은 코드가 아니라 venv 아티팩트였다(uv 3.11 위에 3.13 접목, pip 이 코드가 안 읽는 곳으로). 깨끗한 클론·라이브 트리 양팔 모두 1 failed. 세션마다 13/12/14/1 로 갈리던 원인. 🟥 그리고 codex 가 못 본 형제 경로(not-mapped 분기)를 거버너가 찾았다 — 사이드카는 지목자이지 판정자가 아니라는 계보 그대로"
+
+- date: 2026-09-06
+  agent: general-purpose ×8 (live-eval 진단 · archify U1 재현+PR · qasp 웹QA 레디 점검 · README ko/zh/ja · arXiv 반려 분석[자체 서브에이전트 1 재디스패치] · 논문 문체 감사 · 논문 인용 수정본 · qasp M①M③ 수리) + fh-meta:beginner ×1 (qasp 막별 산출물 콜드리드) + general-purpose ×1 (논문 B 초안) + codex sidecar ×2 (릴리스 델타 cross-family — 1차 stdin 대기로 타임아웃, 2차 성공)
+  invoked_by: FH governor session (Opus 5, 운영자 «오늘 야간 내가 요청한대로 자율주행 이어가줘 … 위임할게 5대정체성 4개 엔진 초록인만큼 잘 써가면서 일해줘»)
+  task: "3.1.0 릴리스 집행 · live-eval 보정 · qasp 웹QA 레디(다음 주 실사용) · arXiv 반려 원인 규명과 논문 재구성 · 랜딩 개편 · archify upstream 기여"
+  context_card: yes (건별 절대경로 · 정본 선독 · 커밋/푸시 경계 · 컨트롤 동반 요구 · 외부 PR 은 금지어휘 목록)
+  outcome: accepted
+  evidence: "릴리스 v3.1.0 npm 라이브(latest=3.1.0) + GitHub 릴리스 + PR #664 머지 · PR #665(verify 재시도) · gh-pages 랜딩 개편 발행(지도 91%→5.6%) · tt-a1i/archify#328 upstream 제출(스위트 1058/1030/0, zip 결정적 재빌드) · qasp PR #269 머지 · 논문 v1.0.1 수정본(인용 21/21 재검증, 원본 해시 불변). cross-family(codex) A급 1건 적발 — CHANGELOG L12 과대주장, 소스 재확인 후 수리, **자력 적발 0**"
+  tokens_subagent: 2100000
+  notes: "🟥 이 날의 최대 발견은 코드가 아니라 **논문**이다 — 반려된 v1.0 의 참고문헌 17건 중 11건이 어긋나 있었다(ID 는 실재, 제목·저자가 다른 논문 것, 한 건은 공간생물학). 그 논문 자신이 «존재 검사 → 출처 검사» 2단계를 방법론으로 제시하고 「허위율 0%」라 보고했다. **방법은 있었고 자기에게 안 돌렸다.** 대조군이 이례적으로 좋다: 같은 심사 창에서 제품명 0회인 자매 논문은 보류 해제. 🟥 부수 교훈 3건이 전부 계기 결함이다 — ① 내 인용 추출기가 References 대신 Changelog 를 긁었다 ② 레인 rc=1 인데 MV 줄만 필터해 보고 실패 3건을 놓쳤다 ③ archify 테스트를 잘못된 cwd 에서 돌려 rc=254 를 «실패»로 읽을 뻔했다. 셋 다 «출력을 좁혀 본 것»이 원인이다"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@chrono-meta/fh-gate",
-  "version": "3.1.0",
+  "version": "3.1.1",
   "description": "FH runtime adapters — run FH governance, skills, and agents via Claude or Codex with machine-parseable gates.",
   "homepage": "https://github.com/chrono-meta/forge-harness#readme",
   "bugs": {

package/plugins/fh-commons/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "fh-commons",
-  "version": "3.1.0",
+  "version": "3.1.1",
   "engines": {
     "claudeCode": ">=1.0.0"
   },

package/plugins/fh-meta/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "fh-meta",
-  "version": "3.1.0",
+  "version": "3.1.1",
   "engines": {
     "claudeCode": ">=1.0.0"
   },

package/plugins/fh-meta/CHANGELOG.md

@@ -1,5 +1,38 @@
 # forge-harness (fh-meta) Changelog
 
+### [3.1.1] — 2026-09-07 — fh-gate Action 의 기본 `level` 이 게이트가 안 받는 값이었다
+
+### 🟥 무엇이 깨져 있었나
+`action.yml` 의 `level` 기본값이 **`'standard'`** 였는데, `scripts/fh-gate.sh:275` 는 `quick` 과
+`full` 만 받고 나머지는 **ARG_ERROR(11)** 로 거절한다. 그리고 `fail-on` 기본 집합에 `ARG_ERROR`
+가 들어 있다. ⇒ **`uses: chrono-meta/forge-harness@v3.1.0` 을 `level:` 없이 쓰면 스텝이 실패했고,
+에러 메시지는 사용자가 설정한 적 없는 인자를 지목했다.**
+
+발행된 3.1.0 타르볼로 실측(소스를 읽은 게 아니다):
+```
+level=standard → rc=11  ERROR: gate level must be 'quick' or 'full' (got: standard)
+level=quick    → rc=12  (DRY_RUN, 정상)
+level=full     → rc=12  (DRY_RUN, 정상)
+```
+
+### 어떻게 잡혔나
+릴리스 마커에 `standpoint: DEGRADED_NOT_RUN — 소비자 install 팔을 안 돌렸다` 라고 적어 두고,
+발행 직후 그 팔을 실제로 돌렸다. **읽어서는 안 나오고 돌려야 나오는 자리**였다.
+레인 28개가 이미 있었지만 전부 **종료코드 매핑**만 봤고 「입력 기본값이 CLI 가 받는 값인가」는
+아무도 안 봤다.
+
+### 고친 것
+- `action.yml` `level` 기본값 `'standard'` → **`'quick'`**, description 도 «다른 값은 안 받는다»로
+- **레인 D0~D2 신설** — `action.yml` 의 기본값과 `fh-gate.sh` 의 허용 집합을 **둘 다 실물에서 추출**해
+  대조한다(어느 쪽도 하드코딩 안 한다 — 그러면 같은 방식으로 또 드리프트한다).
+  D0 = 허용 집합 추출 성공(계기 생사) · D1 = 기본값이 그 안에 있나 · D2 = **컨트롤**(같은 검사가
+  `'standard'` 를 거절하는지 — 아니면 D1 이 아무것도 증명 못 한다). fail-before 실증: 기본값을
+  `standard` 로 되돌리면 **정확히 D1 만** 적색(30 passed · 1 failed), 복원하면 31/0.
+- README 4판의 사용례를 `@v3.1.1` 로
+
+⚠️ **`@v3.1.0` 태그는 그대로 둔다**(태그는 안 움직인다). 그 태그를 쓰는 워크플로는 `level: quick`
+을 명시하면 정상 동작한다.
+
 ### [3.1.0] — 2026-09-06 — QP 플러그인 · `oracle:` 마커 채널 · preprep 도해 굽기 · FH 전체 지도 발행 · fh-gate GitHub Action
 
 ### 왜 minor 인가

package/plugins/fh-qp/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "fh-qp",
-  "version": "3.1.0",
+  "version": "3.1.1",
   "engines": {
     "claudeCode": ">=1.0.0"
   },

package/scripts/com.forge-harness.live-eval.plist

@@ -14,16 +14,16 @@
         <string>/bin/bash</string>
         <string>/path/to/forge-harness/scripts/probe_live_eval.sh</string>
         <!-- 🟥 COST + API-KEY DECISION IS THE OPERATOR'S — this template ships with NO default
-             --subset/--ids limiter, i.e. installing it as-is runs the FULL selected set (12
+             the subset / ids limiter flags limiter, i.e. installing it as-is runs the FULL selected set (12
              probes x 2 calls = 24 live `claude -p` invocations) every night. Read
              scripts/probe_live_eval.sh's own header (§COST) before installing, and add a
-             --subset N or --ids P1,P2 argument here if a nightly full-set run is not what you
+             a subset or ids argument argument here if a nightly full-set run is not what you
              want. This plist is a template, not a recommendation of scope. -->
         <!-- 🟥 reps=3 (2026-09-06). 이 인자가 없으면 reps=1 이고, 그때의 pass_rate 로는 문턱을
              정할 수 없다 — 실측: 유효 런 3 개 재채점에서 12 프로브 중 5 개가 flaky, 같은 코퍼스
              15 분 간격 두 런에서 4 개가 뒤집혔고 관측 pass_rate 는 0.50/0.67/0.67 이었다(단일 rep
              노이즈 폭 > 문턱까지의 거리). 판정은 과반, 분산은 리포트 `reps(pass/ran)` 칸에 남는다.
-             ⚠️ 비용이 3 배다: 24 → **72** 회 `claude -p`. 줄이려면 --subset/--ids 를 같이 줘라. -->
+             ⚠️ 비용이 3 배다: 24 → **72** 회 `claude -p`. 줄이려면 the subset / ids limiter flags 를 같이 줘라. -->
         <string>--reps</string>
         <string>3</string>
     </array>

package/scripts/test_action_yml_lanes.sh

@@ -28,6 +28,69 @@ chk(){ if [ "$1" = "0" ]; then ok "$2"; else no "$2"; fi; }
 
 # ── extract the real case block from the shipped file ────────────────────────────────────────
 MAP="$(awk '/^ *case "\$rc" in/{f=1} f{print} /^ *esac/{if(f){exit}}' "$A" | sed 's/^ *//')"
+
+# ── D. input defaults must be values the CLI accepts (2026-09-06) ────────────────────────────
+# WHY THIS EXISTS. The lanes above test the exit-code MAPPING and nothing else — the header says
+# so. Nobody was checking the other direction: that the values this action sends the CLI are
+# values the CLI takes. v3.1.0 shipped `level` defaulting to 'standard', which `fh-gate.sh` has
+# never accepted; it rejects with ARG_ERROR (11), and ARG_ERROR is in fail-on's own default set.
+# So `uses: chrono-meta/forge-harness@v3.1.0` with no `level:` failed the step, and the error
+# named an argument the user never set. Found by running the PUBLISHED tarball as a consumer
+# would, not by reading either file — which is why the standpoint arm is not optional.
+#
+# Both sides are extracted from the REAL files. A hard-coded expected value here would drift the
+# same way the default did, and this lane would stay green while the shipped file rotted.
+_ACTION_LEVEL_DEFAULT="$(awk '
+  /^  level:/            {inlvl=1; next}
+  inlvl && /^  [a-z-]+:/ {inlvl=0}
+  inlvl && /^ *default:/ {gsub(/^ *default: *|""|\x27/,""); print; exit}
+' "$A")"
+_GATE_SH="$ROOT/scripts/fh-gate.sh"
+if [ ! -f "$_GATE_SH" ]; then
+  no "D0 HARNESS — scripts/fh-gate.sh absent; cannot learn which levels the CLI accepts"
+else
+  # the validation line is the single source of truth for the accepted set
+  _ACCEPTED="$(grep -oE '"\$GATE_LEVEL" != "[a-z]+"' "$_GATE_SH" | grep -oE '"[a-z]+"$' | tr -d '"' | sort -u | tr '\n' ' ')"
+  [ -n "$_ACCEPTED" ] && ok "D0 extracted the CLI's accepted level set from fh-gate.sh: [$_ACCEPTED]" \
+                      || no "D0 could not extract an accepted-level set from fh-gate.sh (instrument dead — the check below would pass vacuously)"
+  case " $_ACCEPTED " in
+    *" $_ACTION_LEVEL_DEFAULT "*) ok "D1 action.yml level default ('$_ACTION_LEVEL_DEFAULT') is a value the CLI accepts" ;;
+    *) no "D1 action.yml level default ('$_ACTION_LEVEL_DEFAULT') is NOT in the CLI's accepted set [$_ACCEPTED] — every step that does not override level: will exit ARG_ERROR" ;;
+  esac
+  # control: the check must reject a value the CLI does not take. If this passes, D1 proves nothing.
+  case " $_ACCEPTED " in
+    *" standard "*) no "D2 control dead — 'standard' appears accepted, so D1 cannot discriminate" ;;
+    *) ok "D2 control — the same test rejects 'standard' (the value v3.1.0 shipped), so D1 discriminates" ;;
+  esac
+
+  # D3/D4 — the SAME check for `backend`, because closing the instance and leaving the class open
+  # is how this defect comes back wearing a different input's name. `backend` is the other
+  # enum-shaped input the action forwards; `model` (free-form, empty = backend default) and
+  # `timeout` (numeric, matches the gate's own 120) have no closed set to drift against, so they
+  # are deliberately not lanes — an assertion with nothing to assert is decoration.
+  _ACTION_BACKEND_DEFAULT="$(awk '
+    /^  backend:/           {inb=1; next}
+    inb && /^  [a-z-]+:/    {inb=0}
+    inb && /^ *default:/    {gsub(/^ *default: *|""|\x27/,""); print; exit}
+  ' "$A")"
+  _ACCEPTED_BE="$(grep -oE "must be '[a-z]+', '[a-z]+', '[a-z]+', or '[a-z]+'" "$_GATE_SH" \
+                   | grep -oE "'[a-z]+'" | tr -d "'" | sort -u | tr '\n' ' ')"
+  if [ -z "$_ACCEPTED_BE" ]; then
+    no "D3 could not extract the accepted backend set from fh-gate.sh (instrument dead — D4 would pass vacuously)"
+  else
+    ok "D3 extracted the CLI's accepted backend set: [$_ACCEPTED_BE]"
+    case " $_ACCEPTED_BE " in
+      *" $_ACTION_BACKEND_DEFAULT "*) ok "D4 action.yml backend default ('$_ACTION_BACKEND_DEFAULT') is a value the CLI accepts" ;;
+      *) no "D4 action.yml backend default ('$_ACTION_BACKEND_DEFAULT') is NOT in [$_ACCEPTED_BE]" ;;
+    esac
+    case " $_ACCEPTED_BE " in
+      *" anthropic "*) no "D5 control dead — a non-existent backend appears accepted" ;;
+      *) ok "D5 control — the same test rejects 'anthropic' (a plausible-but-wrong value), so D4 discriminates" ;;
+    esac
+  fi
+fi
+
+
 [ -n "$MAP" ] || { echo "❌ HARNESS: could not extract the case block from action.yml"; exit 10; }
 printf '%s' "$MAP" | grep -q 'esac' || { echo "❌ HARNESS: extracted block has no esac (truncated)"; exit 10; }