qwerfunch · qwerfunch · Jul 21, 2026 · Jul 20, 2026 · Jul 20, 2026 · Jul 20, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,6 +5,21 @@ All notable changes to Cladding are documented here.
 Format: [Keep a Changelog 1.1.0](https://keepachangelog.com/en/1.1.0/).
 Versioning: [Semantic Versioning 2.0](https://semver.org/spec/v2.0.0.html).
 
+## [Unreleased] — Polyglot gate fidelity + feature-cycle signals
+
+**In one line:** the gate reads and reports non-JavaScript projects honestly, and it now nudges when the feature cycle isn't being driven.
+
+### Gate and toolchain fidelity
+
+- **The check-only formatters list every dirty file, not just the first.** `dart format` (and peers like `dotnet format`) print one changed-file line each; those aren't ESLint-shaped, so the lint stage used to collapse them to a single pathless finding — fix one, re-run, the next appears. Each dirty file now surfaces as its own finding with its path, plus a one-line `fix: run \`dart format .\`` hint.
+- **Python projects run the test suite once per gate, not twice.** On pytest projects the unit stage and the coverage stage each spawned the whole suite (`pytest`, then `coverage run -m pytest`) — roughly doubling gate time. The unit stage now runs the coverage-instrumented suite once and shares it with the coverage stage, the same dedup vitest already had. The zero-executed-tests guard still applies on the shared pytest run, so the dedup can't turn a vacuous run green.
+- **The header-comment convention check understands more languages.** It now recognises `#` comments and Python docstrings (`"""`, `'''`), not only `//` / `/*`.
+
+### Feature-cycle signals
+
+- **A non-blocking nudge when sustained source edits are bound to no feature.** When edits to files that no feature tracks accumulate past a small threshold, the post-edit card surfaces one advisory to start a feature — so the spec-first cycle gets triggered instead of silently skipped. Debounced and once-per-window; it never blocks.
+- **`clad check` flags unenforced projects.** When a project has features not yet done but neither a git hook nor CI wires the gate, `clad check` prints one advisory that the checks run only when asked — with how to wire enforcement. Informational only; it never changes the exit status.
+
 ## [0.9.0] — Project-scoped natural-language onboarding (2026-07-16)
 
 **In one line:** apply Cladding from ordinary conversation without a shell command, while project-local discovery, a read-only preview, an exact approval phrase, and atomic recovery keep onboarding bounded.

diff --git a/README.html b/README.html
@@ -233,7 +233,7 @@ <h1>cladding</h1>
   <p class="badges">
     <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"></a>
     <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"></a>
-    <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests">
+    <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests">
     <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors">
     <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"></a>
   </p>
@@ -548,7 +548,7 @@ <h2>Status</h2>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
         <div style="font-size:11px;color:#64748b;letter-spacing:1.5px;text-transform:uppercase;font-weight:600">tests</div>
-        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">2561<span style="font-size:16px;color:#94a3b8">/2561</span></div>
+        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">2602<span style="font-size:16px;color:#94a3b8">/2602</span></div>
         <div style="font-size:11px;color:#64748b">all pass</div>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
@@ -558,8 +558,8 @@ <h2>Status</h2>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
         <div style="font-size:11px;color:#64748b;letter-spacing:1.5px;text-transform:uppercase;font-weight:600">features</div>
-        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">255</div>
-        <div style="font-size:11px;color:#64748b">252 done · self-spec</div>
+        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">261</div>
+        <div style="font-size:11px;color:#64748b">258 done · self-spec</div>
       </td>
     </tr>
   </table>

diff --git a/README.ja.md b/README.ja.md
@@ -12,7 +12,7 @@
 <p align="center">
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"/></a>
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"/></a>
-  <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests"/>
+  <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests"/>
   <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors"/>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"/></a>
 </p>
@@ -339,7 +339,7 @@ clad update              # 3. プロジェクト接続と派生状態を更新
 
 | Version | 準拠レベル | Tests | Gate | Features |
 |---|---|---|---|---|
-| v0.9.0（2026-07） | L4 · [自己申告](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2561 / 2561 | 15 段階 · 41 detectors | 255（252 done） |
+| v0.9.0（2026-07） | L4 · [自己申告](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2602 / 2602 | 15 段階 · 41 detectors | 261（258 done） |
 
 <sub>236 test files · capability 6 個 · カバレッジ低下は COVERAGE_DROP detector がブロック</sub>
 

diff --git a/README.ko.html b/README.ko.html
@@ -275,7 +275,7 @@ <h1>cladding</h1>
   <p class="badges">
     <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"></a>
     <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"></a>
-    <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests">
+    <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests">
     <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors">
     <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"></a>
   </p>
@@ -584,7 +584,7 @@ <h2>Status</h2>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
         <div style="font-size:11px;color:#64748b;letter-spacing:1.5px;text-transform:uppercase;font-weight:600">tests</div>
-        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">2561<span style="font-size:16px;color:#94a3b8">/2561</span></div>
+        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">2602<span style="font-size:16px;color:#94a3b8">/2602</span></div>
         <div style="font-size:11px;color:#64748b">all pass</div>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
@@ -594,8 +594,8 @@ <h2>Status</h2>
       </td>
       <td style="text-align:center;width:140px;background:#f8fafc;padding:18px 10px;border-radius:8px;border:none">
         <div style="font-size:11px;color:#64748b;letter-spacing:1.5px;text-transform:uppercase;font-weight:600">features</div>
-        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">255</div>
-        <div style="font-size:11px;color:#64748b">252 done · 자기 스펙</div>
+        <div style="font-size:24px;font-weight:800;color:#0f172a;margin:8px 0;letter-spacing:-0.5px">261</div>
+        <div style="font-size:11px;color:#64748b">258 done · 자기 스펙</div>
       </td>
     </tr>
   </table>

diff --git a/README.ko.md b/README.ko.md
@@ -12,7 +12,7 @@
 <p align="center">
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"/></a>
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"/></a>
-  <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests"/>
+  <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests"/>
   <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors"/>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"/></a>
 </p>
@@ -338,7 +338,7 @@ clad update              # 3. 프로젝트 연결과 파생 데이터를 함께
 
 | version | 준수 등급 | tests | gate | features |
 |---|---|---|---|---|
-| v0.9.0 · 2026-07 | L4 · [L0–L4 중 최고 · 자가 선언](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2561 / 2561 · all pass | 15 단계 · 41 detectors | 255 · 252 done · 자기 스펙 |
+| v0.9.0 · 2026-07 | L4 · [L0–L4 중 최고 · 자가 선언](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2602 / 2602 · all pass | 15 단계 · 41 detectors | 261 · 258 done · 자기 스펙 |
 
 <sub>236 test files · capability 6개 · coverage는 COVERAGE_DROP detector가 하락 차단</sub>
 

diff --git a/README.md b/README.md
@@ -12,7 +12,7 @@
 <p align="center">
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"/></a>
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"/></a>
-  <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests"/>
+  <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests"/>
   <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors"/>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"/></a>
 </p>
@@ -352,7 +352,7 @@ Reconcile the drift the update flagged.
 
 | Version | Conformance | Tests | Gate | Features |
 |---|---|---|---|---|
-| v0.9.0 (2026-07) | L4 · [self-declared](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2561 / 2561 | 15 stages · 41 detectors | 255 (252 done) |
+| v0.9.0 (2026-07) | L4 · [self-declared](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2602 / 2602 | 15 stages · 41 detectors | 261 (258 done) |
 
 <sub>236 test files · 6 capabilities · coverage drop blocked by the COVERAGE_DROP detector</sub>
 

diff --git a/README.zh.md b/README.zh.md
@@ -12,7 +12,7 @@
 <p align="center">
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/ironclad-L4%20conformant-brightgreen" alt="ironclad"/></a>
   <a href="https://github.com/qwerfunch/ironclad"><img src="https://img.shields.io/badge/spec-v0.0.23-blue" alt="spec"/></a>
-  <img src="https://img.shields.io/badge/tests-2561%2F2561-brightgreen" alt="tests"/>
+  <img src="https://img.shields.io/badge/tests-2602%2F2602-brightgreen" alt="tests"/>
   <img src="https://img.shields.io/badge/detectors-41-brightgreen" alt="detectors"/>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-lightgrey" alt="license"/></a>
 </p>
@@ -335,7 +335,7 @@ clad update              # 3. 刷新项目连接和派生状态
 
 | 版本 | 一致性 | Tests | Gate | Features |
 |---|---|---|---|---|
-| v0.9.0（2026-07） | L4 · [自我声明](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2561 / 2561 | 15 阶段 · 41 检测器 | 255（252 done） |
+| v0.9.0（2026-07） | L4 · [自我声明](https://github.com/qwerfunch/ironclad/blob/main/GOVERNANCE.md) | 2602 / 2602 | 15 阶段 · 41 检测器 | 261（258 done） |
 
 <sub>236 个测试文件 · 6 项 capability · 覆盖率下降由 COVERAGE_DROP 检测器拦下</sub>
 

diff --git a/docs/ab-evaluation/case-existing-adoption.md b/docs/ab-evaluation/case-existing-adoption.md
@@ -51,8 +51,8 @@ No spec, no architecture invariants — just code on the existing tree.
 | Test files | 1 | 1 | +0 |
 | Test LoC | 14 | 14 | +0 |
 | Test cases | 1 | 1 | +0 |
-| Total chars (artifacts + code) | 11212 | 3864 | +7348 |
-| Estimated tokens | 2805 | 967 | +1838 |
+| Total chars (artifacts + code) | 11104 | 3864 | +7240 |
+| Estimated tokens | 2778 | 967 | +1811 |
 
 **Detector outcomes** (META_INTEGRITY + HARDCODED_SECRET excluded — toolchain-only checks):
 
@@ -90,8 +90,8 @@ B (Vanilla)  — errors: 1  warns: 3  infos: 28
 | Test files | 2 | 2 | +0 |
 | Test LoC | 27 | 27 | +0 |
 | Test cases | 3 | 3 | +0 |
-| Total chars (artifacts + code) | 12732 | 5115 | +7617 |
-| Estimated tokens | 3185 | 1280 | +1905 |
+| Total chars (artifacts + code) | 12624 | 5115 | +7509 |
+| Estimated tokens | 3158 | 1280 | +1878 |
 
 **Detector outcomes** (META_INTEGRITY + HARDCODED_SECRET excluded — toolchain-only checks):
 
@@ -111,7 +111,7 @@ B (Vanilla)  — errors: 1  warns: 3  infos: 28
 - **Spec ↔ code traceability**: cladding emits 1 feature(s), 2 AC(s), 1 scenario(s), 3 capability(s); vanilla has 0 of each.
 - **Architecture enforcement**: cladding declares 3 layer(s) with 0 forbidden-import rule(s); vanilla has 0.
 - **Detector behavior**: cladding-managed tree → 1 error(s) / 1 warn(s) / 11 info(s). Vanilla tree → 1 / 3 / 28. The detectors that gate against spec (REFERENCE_INTEGRITY, MISSING_IMPLEMENTATION, ARCHITECTURE_FROM_SPEC, CAPABILITIES_FEATURE_MAPPING) need cladding's artifacts to evaluate — without them they silently pass. The "0 errors on vanilla" therefore is **absence of signal**, not absence of drift.
-- **Token cost**: cladding's cumulative artifact + code consumes ~3185 tokens vs vanilla's ~1280 (heuristic chars/4) — Δ ≈ 1905 tokens, the price of structure.
+- **Token cost**: cladding's cumulative artifact + code consumes ~3158 tokens vs vanilla's ~1280 (heuristic chars/4) — Δ ≈ 1878 tokens, the price of structure.
 - **Code surface**: vanilla writes 9 source file(s) / 144 LoC + 2 test file(s) / 3 test case(s); cladding writes 9 / 128 + 2 / 3. (Vanilla front-loads code, cladding front-loads spec — both converge by M2.)
 
 ## Outcome Quality (F-ba2e05)

diff --git a/docs/ab-evaluation/case-payment-saas.md b/docs/ab-evaluation/case-payment-saas.md
@@ -50,8 +50,8 @@ no spec, no scenarios, no architecture invariants.
 | Test files | 0 | 1 | -1 |
 | Test LoC | 0 | 20 | -20 |
 | Test cases | 0 | 2 | -2 |
-| Total chars (artifacts + code) | 6708 | 3463 | +3245 |
-| Estimated tokens | 1679 | 867 | +812 |
+| Total chars (artifacts + code) | 6600 | 3463 | +3137 |
+| Estimated tokens | 1652 | 867 | +785 |
 
 **Detector outcomes** (META_INTEGRITY + HARDCODED_SECRET excluded — toolchain-only checks):
 
@@ -89,8 +89,8 @@ B (Vanilla)  — errors: 1  warns: 3  infos: 28
 | Test files | 1 | 2 | -1 |
 | Test LoC | 8 | 33 | -25 |
 | Test cases | 1 | 4 | -3 |
-| Total chars (artifacts + code) | 7872 | 5592 | +2280 |
-| Estimated tokens | 1970 | 1399 | +571 |
+| Total chars (artifacts + code) | 7764 | 5592 | +2172 |
+| Estimated tokens | 1943 | 1399 | +544 |
 
 **Detector outcomes** (META_INTEGRITY + HARDCODED_SECRET excluded — toolchain-only checks):
 
@@ -110,7 +110,7 @@ B (Vanilla)  — errors: 1  warns: 3  infos: 28
 - **Spec ↔ code traceability**: cladding emits 1 feature(s), 2 AC(s), 2 scenario(s), 3 capability(s); vanilla has 0 of each.
 - **Architecture enforcement**: cladding declares 3 layer(s) with 2 forbidden-import rule(s); vanilla has 0.
 - **Detector behavior**: cladding-managed tree → 1 error(s) / 3 warn(s) / 15 info(s). Vanilla tree → 1 / 3 / 28. The detectors that gate against spec (REFERENCE_INTEGRITY, MISSING_IMPLEMENTATION, ARCHITECTURE_FROM_SPEC, CAPABILITIES_FEATURE_MAPPING) need cladding's artifacts to evaluate — without them they silently pass. The "0 errors on vanilla" therefore is **absence of signal**, not absence of drift.
-- **Token cost**: cladding's cumulative artifact + code consumes ~1970 tokens vs vanilla's ~1399 (heuristic chars/4) — Δ ≈ 571 tokens, the price of structure.
+- **Token cost**: cladding's cumulative artifact + code consumes ~1943 tokens vs vanilla's ~1399 (heuristic chars/4) — Δ ≈ 544 tokens, the price of structure.
 - **Code surface**: vanilla writes 5 source file(s) / 126 LoC + 2 test file(s) / 4 test case(s); cladding writes 1 / 11 + 1 / 1. (Vanilla front-loads code, cladding front-loads spec — both converge by M2.)
 
 ## Outcome Quality (F-ba2e05)

diff --git a/plugins/antigravity/skills/init/SKILL.md b/plugins/antigravity/skills/init/SKILL.md
@@ -19,7 +19,7 @@ If the current user message itself matches `APPLY CLADDING XXXXXX`, do not call
 3. Read the returned prompt and observations. Draft the structured object required by `clad_init` using the current host model, then call `clad_stage_init` with the preparation token and that draft. Staging validates the draft and stores only ignored runtime state under `.cladding/host/`; it does not modify authored project files.
 4. Show `plannedChanges`, the staged `confirmationQuestion`, and its one-time `approvalChallenge` to the user, then stop and wait for a separate reply. The original initialization request is not confirmation.
 5. Only when the user's separate reply exactly matches `approvalChallenge`, call `clad_init` with the entire reply verbatim as `confirmation`—never strip the `APPLY CLADDING` prefix. Include the draft and token when the host retained them; process-per-turn hosts may omit both because Cladding resolves the exact staged draft from its short-lived machine-local cache. Questions, paraphrases, and generic acknowledgements are not approval.
-6. If `nextQuestion` is present, show it verbatim and end the assistant turn immediately. Never answer it, infer an answer, or call either clarify tool during the initialization-approval turn. Continue with the Cladding clarify workflow only after a new user message supplies the answer. If no question remains, report that ordinary development can begin.
+6. If `nextQuestion` is present, show it verbatim and end the assistant turn immediately. Never answer it, infer an answer, or call either clarify tool during the initialization-approval turn. Continue with the Cladding clarify workflow only after a new user message supplies the answer. If no question remains, report that onboarding is complete and that the next step is to author the first feature's spec — its acceptance criteria and the files it will cover — before writing any code; the feature cycle starts there, not with scaffolding.
 
 Do not run `clad init` in a shell from an AI-host onboarding session. Do not use MCP sampling. If the MCP tools are absent, tell the user to run `clad setup`, restart the AI host in the project directory, and stop without writing project files.