@fitlab-ai/agent-infra 0.8.3 → 0.8.4

package/templates/.agents/skills/review-analysis/reference/output-templates.en.md

@@ -27,7 +27,7 @@ Task {task-id} requirement analysis review completed. Verdict: approved.
 
 Next step - write the technical plan:
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -43,12 +43,12 @@ Task {task-id} requirement analysis review completed. Verdict: approved.
 
 Next step - revise analysis before continuing (recommended):
   - Claude Code / OpenCode: /analyze-task {task-ref}
-  - Gemini CLI: /agent-infra:analyze-task {task-ref}
+  - Gemini CLI: /{{project}}:analyze-task {task-ref}
   - Codex CLI: $analyze-task {task-ref}
 
 Or proceed directly to planning:
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -64,7 +64,7 @@ Task {task-id} requirement analysis review completed. Verdict: changes requested
 
 Next step - revise requirement analysis:
   - Claude Code / OpenCode: /analyze-task {task-ref}
-  - Gemini CLI: /agent-infra:analyze-task {task-ref}
+  - Gemini CLI: /{{project}}:analyze-task {task-ref}
   - Codex CLI: $analyze-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -80,7 +80,7 @@ Task {task-id} requirement analysis review completed. Verdict: rejected, fresh a
 
 Next step - re-analyze:
   - Claude Code / OpenCode: /analyze-task {task-ref}
-  - Gemini CLI: /agent-infra:analyze-task {task-ref}
+  - Gemini CLI: /{{project}}:analyze-task {task-ref}
   - Codex CLI: $analyze-task {task-ref}
 
 [When manual-validation > 0, append:]

package/templates/.agents/skills/review-analysis/reference/output-templates.zh-CN.md

@@ -27,7 +27,7 @@
 
 下一步 - 编写技术方案：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -43,12 +43,12 @@
 
 下一步 - 修订分析后继续（推荐）：
   - Claude Code / OpenCode：/analyze-task {task-ref}
-  - Gemini CLI：/agent-infra:analyze-task {task-ref}
+  - Gemini CLI：/{{project}}:analyze-task {task-ref}
   - Codex CLI：$analyze-task {task-ref}
 
 或直接进入方案设计：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -64,7 +64,7 @@
 
 下一步 - 修订需求分析：
   - Claude Code / OpenCode：/analyze-task {task-ref}
-  - Gemini CLI：/agent-infra:analyze-task {task-ref}
+  - Gemini CLI：/{{project}}:analyze-task {task-ref}
   - Codex CLI：$analyze-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -80,7 +80,7 @@
 
 下一步 - 重新分析：
   - Claude Code / OpenCode：/analyze-task {task-ref}
-  - Gemini CLI：/agent-infra:analyze-task {task-ref}
+  - Gemini CLI：/{{project}}:analyze-task {task-ref}
   - Codex CLI：$analyze-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]

package/templates/.agents/skills/review-code/SKILL.en.md

@@ -74,9 +74,10 @@ Read the highest-round code artifact and, if present, the highest-round fix arti
 ### 4. Perform the Review
 
 Follow `.agents/workflows/feature-development.yaml` and inspect the full change context:
-- `git diff --binary HEAD -- <post-review-globs>` for tracked changes
+- Capture `R=$(git rev-parse HEAD)` exactly once; reuse that R for this round's report, fingerprint, and task review fact, and do not re-read HEAD later as a substitute
+- `git diff --binary "$R" -- <post-review-globs>` for tracked changes
 - `git ls-files -o --exclude-standard -z -- <post-review-globs>` for untracked new files
-- `node .agents/scripts/review-diff-fingerprint.js worktree HEAD` for the reviewed diff fingerprint; write it into the report
+- `node .agents/scripts/review-diff-fingerprint.js worktree "$R"` for the reviewed diff fingerprint; write it into the report
 
 > Detailed review criteria, severity rules, and reviewer expectations live in `reference/review-criteria.md`. Read `reference/review-criteria.md` before reviewing.
 > Test review gate: when `git diff` touches test files, read `.agents/rules/testing-discipline.md` first and check it item by item, especially "do not add negative assertions when a positive assertion already covers the behavior".
@@ -92,10 +93,13 @@ Create `.agents/workspace/active/{task-id}/{review-artifact}`.
 Get the current time:
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
-Update task.md and append:
+Update task.md:
+- When this round's `Overall Verdict` / `总体结论` is `Approved` / `通过`, write `last_reviewed_commit: {R}`; the presence of a worktree diff does not affect this write, and repeating the same R is idempotent
+- When this round is not Approved, preserve the existing `last_reviewed_commit`; do not advance or clear it
+- Append:
 `- {YYYY-MM-DD HH:mm:ss±HH:MM} — **Review Code (Round {N})** by {agent} — Verdict: {Approved/Changes Requested/Rejected}, blockers: {n}, major: {n}, minor: {n}, Manual-validation: {n} → {artifact-filename}`
 
 Always include the `Manual-validation: {n}` field in the done log, including when it is 0.

package/templates/.agents/skills/review-code/SKILL.zh-CN.md

@@ -73,9 +73,10 @@ tail .agents/workspace/active/{task-id}/task.md
 ### 4. 执行审查
 
 遵循 `.agents/workflows/feature-development.yaml`，并同时检查完整变更上下文：
-- `git diff --binary HEAD -- <post-review-globs>` 覆盖已跟踪变更
+- 一次性记录 `R=$(git rev-parse HEAD)`；本轮报告、指纹和任务审查事实都复用该 R，禁止稍后重新读取 HEAD 代替
+- `git diff --binary "$R" -- <post-review-globs>` 覆盖已跟踪变更
 - `git ls-files -o --exclude-standard -z -- <post-review-globs>` 覆盖未跟踪新文件
-- `node .agents/scripts/review-diff-fingerprint.js worktree HEAD` 生成审查差异指纹，并写入报告
+- `node .agents/scripts/review-diff-fingerprint.js worktree "$R"` 生成审查差异指纹，并写入报告
 
 > 详细审查标准、严重程度划分和 reviewer 关注点见 `reference/review-criteria.md`。执行此步骤前先读取 `reference/review-criteria.md`。
 > 测试审查硬门禁：当 `git diff` 触及测试文件时，必须先读取 `.agents/rules/testing-discipline.md` 并逐条核对（尤其"正向已覆盖时不应再加反向断言"）。
@@ -91,13 +92,15 @@ tail .agents/workspace/active/{task-id}/task.md
 获取当前时间：
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
 - `current_step`：code-review
 - `assigned_to`：{当前代理}
 - `updated_at`：{当前时间}
 - `agent_infra_version`：按 `.agents/rules/version-stamp.md` 取值
+- 若本轮 `总体结论` / `Overall Verdict` 为 `通过` / `Approved`，写入 `last_reviewed_commit: {R}`；是否存在工作区 diff 不影响写入，重复写同一 R 幂等
+- 若本轮结论不是 Approved，保留既有 `last_reviewed_commit`，不得推进或清空
 - 追加：
 `- {YYYY-MM-DD HH:mm:ss±HH:MM} — **Review Code (Round {N})** by {agent} — Verdict: {Approved/Changes Requested/Rejected}, blockers: {n}, major: {n}, minor: {n}, Manual-validation: {n} → {artifact-filename}`
 

package/templates/.agents/skills/review-code/config/verify.en.json

@@ -36,6 +36,7 @@
       ],
       "freshness_minutes": 30
     },
+    "review-fact": {},
     "activity-log": {
       "expected_action_pattern": "(Review Code|Code Review) \\(Round \\d+\\)",
       "freshness_minutes": 30

package/templates/.agents/skills/review-code/config/verify.zh-CN.json

@@ -36,6 +36,7 @@
       ],
       "freshness_minutes": 30
     },
+    "review-fact": {},
     "activity-log": {
       "expected_action_pattern": "(Review Code|Code Review) \\(Round \\d+\\)",
       "freshness_minutes": 30

package/templates/.agents/skills/review-code/reference/output-templates.en.md

@@ -29,7 +29,7 @@ Task {task-id} review completed. Verdict: approved.
 
 Next step - commit the code:
   - Claude Code / OpenCode: /commit
-  - Gemini CLI: /agent-infra:commit
+  - Gemini CLI: /{{project}}:commit
   - Codex CLI: $commit
 
 [When manual-validation > 0, append this final line:]
@@ -45,12 +45,12 @@ Task {task-id} review completed. Verdict: approved.
 
 Next step - fix before commit (recommended):
   - Claude Code / OpenCode: /code-task {task-ref}
-  - Gemini CLI: /agent-infra:code-task {task-ref}
+  - Gemini CLI: /{{project}}:code-task {task-ref}
   - Codex CLI: $code-task {task-ref}
 
 Or commit directly (skip fix):
   - Claude Code / OpenCode: /commit
-  - Gemini CLI: /agent-infra:commit
+  - Gemini CLI: /{{project}}:commit
   - Codex CLI: $commit
 
 [When manual-validation > 0, append this final line:]
@@ -66,7 +66,7 @@ Task {task-id} review completed. Verdict: changes requested.
 
 Next step - fix the findings:
   - Claude Code / OpenCode: /code-task {task-ref}
-  - Gemini CLI: /agent-infra:code-task {task-ref}
+  - Gemini CLI: /{{project}}:code-task {task-ref}
   - Codex CLI: $code-task {task-ref}
 
 [When manual-validation > 0, append this final line:]
@@ -82,7 +82,7 @@ Task {task-id} review completed. Verdict: rejected, re-design the technical plan
 
 Next step - re-design the technical plan:
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 > Note: Rejected means the implementation direction needs to be reworked end-to-end, not patched locally. `code-task/scripts/detect-mode.js` branch #7 refuses a direct `/code-task` and requires a fresh plan first.

package/templates/.agents/skills/review-code/reference/output-templates.zh-CN.md

@@ -29,7 +29,7 @@
 
 下一步 - 提交代码：
   - Claude Code / OpenCode：/commit
-  - Gemini CLI：/agent-infra:commit
+  - Gemini CLI：/{{project}}:commit
   - Codex CLI：$commit
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -45,12 +45,12 @@
 
 下一步 - 修复问题后提交（推荐）：
   - Claude Code / OpenCode：/code-task {task-ref}
-  - Gemini CLI：/agent-infra:code-task {task-ref}
+  - Gemini CLI：/{{project}}:code-task {task-ref}
   - Codex CLI：$code-task {task-ref}
 
 或直接提交（跳过修复）：
   - Claude Code / OpenCode：/commit
-  - Gemini CLI：/agent-infra:commit
+  - Gemini CLI：/{{project}}:commit
   - Codex CLI：$commit
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -66,7 +66,7 @@
 
 下一步 - 修复问题：
   - Claude Code / OpenCode：/code-task {task-ref}
-  - Gemini CLI：/agent-infra:code-task {task-ref}
+  - Gemini CLI：/{{project}}:code-task {task-ref}
   - Codex CLI：$code-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -82,7 +82,7 @@
 
 下一步 - 重新设计技术方案：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 > 注意：Rejected 表示实现方向需要整体重做，不是局部修复。`code-task/scripts/detect-mode.js` 分支 #7 会拒绝直接 `/code-task`，要求先重新方案设计。

package/templates/.agents/skills/review-code/reference/report-template.en.md

@@ -21,8 +21,8 @@ Use this template when writing `review-code.md` or `review-code-r{N}.md`.
 - **Reviewer**: {reviewer-name}
 - **Review Time**: {timestamp}
 - **Scope**: {file-count and major modules}
-- **Review Baseline Commit**: {raw git rev-parse HEAD} (baseline for the complete-task post-review commit gate; see `.agents/rules/review-handshake.md`)
-- **Reviewed Diff Fingerprint**: {raw node .agents/scripts/review-diff-fingerprint.js worktree HEAD}
+- **Review Baseline Commit**: {raw R captured once for this round} (baseline for the complete-task post-review commit gate; see `.agents/rules/review-handshake.md`)
+- **Reviewed Diff Fingerprint**: {raw node .agents/scripts/review-diff-fingerprint.js worktree "$R"}
 - **Overall Verdict**: {Approved / Changes Requested / Rejected} (pick exactly one; combined phrases will fail the verify gate)
 - **Findings (AI-actionable)**: 0 blockers, 0 majors, 0 minors / **Manual validation**: 0
 

package/templates/.agents/skills/review-code/reference/report-template.zh-CN.md

@@ -21,8 +21,8 @@
 - **审查者**：{reviewer-name}
 - **审查时间**：{timestamp}
 - **审查范围**：{file-count and major modules}
-- **审查基线提交**：{git rev-parse HEAD 原文}（complete-task 的 post-review commit 门禁基线；详见 `.agents/rules/review-handshake.md`）
-- **审查差异指纹**：{node .agents/scripts/review-diff-fingerprint.js worktree HEAD 原文}
+- **审查基线提交**：{本轮一次性捕获的 R 原文}（complete-task 的 post-review commit 门禁基线；详见 `.agents/rules/review-handshake.md`）
+- **审查差异指纹**：{node .agents/scripts/review-diff-fingerprint.js worktree "$R" 原文}
 - **总体结论**：{通过 / 需要修改 / 拒绝}（恰取一个；禁止写组合短语，否则 verify gate 失败）
 - **发现（AI 可处理）**：0 阻塞项，0 主要，0 次要 / **人工校验**：0
 

package/templates/.agents/skills/review-plan/SKILL.en.md

@@ -73,7 +73,7 @@ Create `.agents/workspace/active/{task-id}/{review-artifact}`.
 Get the current time:
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
 Set `current_step` to `technical-design-review`, refresh task metadata, and append:

package/templates/.agents/skills/review-plan/SKILL.zh-CN.md

@@ -79,7 +79,7 @@ tail .agents/workspace/active/{task-id}/task.md
 获取当前时间：
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
 更新 task.md：

package/templates/.agents/skills/review-plan/reference/output-templates.en.md

@@ -27,7 +27,7 @@ Task {task-id} technical plan review completed. Verdict: approved.
 
 Next step - write code:
   - Claude Code / OpenCode: /code-task {task-ref}
-  - Gemini CLI: /agent-infra:code-task {task-ref}
+  - Gemini CLI: /{{project}}:code-task {task-ref}
   - Codex CLI: $code-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -43,12 +43,12 @@ Task {task-id} technical plan review completed. Verdict: approved.
 
 Next step - revise plan before coding (recommended):
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 Or proceed directly to coding:
   - Claude Code / OpenCode: /code-task {task-ref}
-  - Gemini CLI: /agent-infra:code-task {task-ref}
+  - Gemini CLI: /{{project}}:code-task {task-ref}
   - Codex CLI: $code-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -64,7 +64,7 @@ Task {task-id} technical plan review completed. Verdict: changes requested.
 
 Next step - revise technical plan:
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 [When manual-validation > 0, append:]
@@ -80,7 +80,7 @@ Task {task-id} technical plan review completed. Verdict: rejected, redesign requ
 
 Next step - redesign:
   - Claude Code / OpenCode: /plan-task {task-ref}
-  - Gemini CLI: /agent-infra:plan-task {task-ref}
+  - Gemini CLI: /{{project}}:plan-task {task-ref}
   - Codex CLI: $plan-task {task-ref}
 
 [When manual-validation > 0, append:]

package/templates/.agents/skills/review-plan/reference/output-templates.zh-CN.md

@@ -27,7 +27,7 @@
 
 下一步 - 编写代码：
   - Claude Code / OpenCode：/code-task {task-ref}
-  - Gemini CLI：/agent-infra:code-task {task-ref}
+  - Gemini CLI：/{{project}}:code-task {task-ref}
   - Codex CLI：$code-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -43,12 +43,12 @@
 
 下一步 - 修订方案后编码（推荐）：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 或直接进入编码：
   - Claude Code / OpenCode：/code-task {task-ref}
-  - Gemini CLI：/agent-infra:code-task {task-ref}
+  - Gemini CLI：/{{project}}:code-task {task-ref}
   - Codex CLI：$code-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -64,7 +64,7 @@
 
 下一步 - 修订技术方案：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]
@@ -80,7 +80,7 @@
 
 下一步 - 重新设计：
   - Claude Code / OpenCode：/plan-task {task-ref}
-  - Gemini CLI：/agent-infra:plan-task {task-ref}
+  - Gemini CLI：/{{project}}:plan-task {task-ref}
   - Codex CLI：$plan-task {task-ref}
 
 [当 manual-validation > 0 时，在最后附加一行：]

package/templates/.agents/skills/test/SKILL.zh-CN.md

@@ -11,16 +11,27 @@ description: >
 
 ## 1. 编译 / 类型检查
 
-本项目由 Node.js CLI 和模板文件组成，无需编译。跳过此步骤。
+```bash
+# TODO：替换为项目实际的编译命令
+# npx tsc --noEmit       (TypeScript)
+# mvn compile             (Maven)
+# go build ./...          (Go)
+# make build              (通用)
+```
+
+确认没有编译错误。
 
 ## 2. 运行单元测试（按层级选择）
 
-本项目把测试分为三层，按场景选择运行命令；新增测试文件默认归入 **full**，确认足够快且足够核心后，再上调到 core 或 smoke。
+三层测试是可选的反馈速度优化；如果测试套件较小，所有层级都可以映射到同一个完整测试命令。
 
 ### smoke（目标 <5s）
 
 ```bash
-npm run test:smoke
+# TODO：替换为项目的 smoke 子集命令
+# npm run test:smoke       (Node.js)
+# pytest -m "not slow"     (Python)
+# go test -short ./...     (Go)
 ```
 
 适用场景：
@@ -31,7 +42,10 @@ npm run test:smoke
 ### core（目标 <15s）
 
 ```bash
-npm run test:core
+# TODO：替换为项目的 core 子集命令
+# npm run test:core        (Node.js)
+# pytest -m "not contract" (Python)
+# go test ./...            (Go)
 ```
 
 适用场景：
@@ -39,18 +53,22 @@ npm run test:core
 - 写 code.md / code-r{N}.md 报告前的最终验证
 - 推送 PR 前的本地把关
 
-### full（目标 <60s）
+### full（完整测试套件）
 
 ```bash
-npm test
+# TODO：替换为项目的完整测试命令
+# npm test                 (Node.js)
+# mvn test                 (Maven)
+# pytest                   (Python)
+# go test ./...            (Go)
 ```
 
 适用场景：
 - release / tag 前
-- CI（unit-tests.yml）
+- CI
 - main 合并前的最终把关
 
-full 层运行全部项目测试。`npm test` 使用通配匹配项目测试文件，**新增的测试文件会自动归入 full**，这是安全网。
+如果项目尚未分层，smoke / core / full 可以全部使用完整测试命令；分层不是使用协作工作流的前置条件。
 
 ## 3. 输出结果
 

package/templates/.agents/skills/update-agent-infra/SKILL.zh-CN.md

@@ -133,7 +133,7 @@ node .agents/skills/update-agent-infra/scripts/sync-templates.js
 ```
 下一步 - 提交代码：
   - Claude Code / OpenCode：/commit
-  - Gemini CLI：/agent-infra:commit
+  - Gemini CLI：/{{project}}:commit
   - Codex CLI：$commit
 ```
 

package/templates/.agents/skills/update-agent-infra/scripts/sync-templates.js

@@ -73,6 +73,7 @@ const DEFAULTS = {
       "**/test-integration.*",
       "**/test.*",
       "**/upgrade-dependency.*",
+      ".agents/rules/testing-discipline.*",
       ".agents/skills/post-release/SKILL.*",
       ".agents/skills/release/SKILL.*",
       ".agents/skills/test-integration/SKILL.*",
@@ -235,7 +236,7 @@ function parseSkillFrontmatter(filePath) {
     if (!pair) continue;
 
     const [, key, rawValue] = pair;
-    if (rawValue === '>') {
+    if (rawValue === '>' || rawValue === '|') {
       const block = [];
       for (let offset = index + 1; offset < lines.length; offset += 1) {
         const nextLine = lines[offset];
@@ -244,7 +245,7 @@ function parseSkillFrontmatter(filePath) {
         block.push(nextLine.trim());
         index = offset;
       }
-      result[key] = block.join(' ').trim();
+      result[key] = block.join(rawValue === '|' ? '\n' : ' ').trim();
       continue;
     }
 
@@ -286,7 +287,8 @@ function detectCustomSkills(projectRoot, templateSkillNames) {
         dirName: entry.name,
         name: meta.name || entry.name,
         description: meta.description || '',
-        args: meta.args || null
+        args: meta.args || null,
+        disableModelInvocation: meta['disable-model-invocation'] === 'true'
       };
     })
     .filter(Boolean)
@@ -501,14 +503,35 @@ function cleanStaleSyncedFiles(projectRoot, syncedSkills, report) {
   }
 }
 
+function formatYamlMetadata(key, value) {
+  if (!value.includes('\n')) {
+    return [`${key}: ${JSON.stringify(value)}`];
+  }
+
+  return [`${key}: |-`, ...value.split('\n').map((line) => `  ${line}`)];
+}
+
+function formatTomlMetadata(key, value) {
+  if (!value.includes('\n')) {
+    return `${key} = ${JSON.stringify(value)}`;
+  }
+
+  const lines = value.split('\n').map((line) => JSON.stringify(line).slice(1, -1));
+  return `${key} = """${lines.join('\n')}"""`;
+}
+
 function generateClaudeCommand(skill, lang) {
   const isZhCN = lang === 'zh-CN';
-  const lines = ['---', `description: ${JSON.stringify(skill.description)}`];
+  const lines = ['---', ...formatYamlMetadata('description', skill.description)];
 
   if (skill.args) {
     lines.push(`usage: ${JSON.stringify(`/${skill.dirName} ${skill.args}`)}`);
   }
 
+  if (skill.disableModelInvocation) {
+    lines.push('disable-model-invocation: true');
+  }
+
   lines.push('---', '');
   lines.push(
     isZhCN
@@ -539,7 +562,7 @@ function generateGeminiCommand(skill, lang) {
   promptLines.push(isZhCN ? '严格按照技能中定义的所有步骤执行。' : 'Follow all steps defined in the skill exactly.');
 
   return [
-    `description = ${JSON.stringify(skill.description)}`,
+    formatTomlMetadata('description', skill.description),
     'prompt = """',
     ...promptLines,
     '"""'
@@ -550,7 +573,7 @@ function generateOpenCodeCommand(skill, lang) {
   const isZhCN = lang === 'zh-CN';
   const lines = [
     '---',
-    `description: ${JSON.stringify(skill.description)}`,
+    ...formatYamlMetadata('description', skill.description),
     'agent: general',
     'subtask: false',
     '---',

package/templates/.agents/skills/watch-pr/SKILL.en.md

@@ -67,7 +67,7 @@ When self-heal hits the cap, the failure is non-code, the run id is unlocatable,
 Get the current time:
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
 Update `.agents/workspace/active/{task-id}/task.md`:
@@ -109,7 +109,7 @@ Output per scenario:
   ```
   Next step - Complete and archive the task:
     - Claude Code / OpenCode: /complete-task {task-ref}
-    - Gemini CLI: /agent-infra:complete-task {task-ref}
+    - Gemini CLI: /{{project}}:complete-task {task-ref}
     - Codex CLI: $complete-task {task-ref}
   ```
 

package/templates/.agents/skills/watch-pr/SKILL.zh-CN.md

@@ -67,7 +67,7 @@ description: >
 获取当前时间：
 
 ```bash
-date "+%Y-%m-%d %H:%M:%S%:z"
+date "+%Y-%m-%d %H:%M:%S%z" | sed 's/\([+-][0-9][0-9]\)\([0-9][0-9]\)$/\1:\2/'
 ```
 
 更新 `.agents/workspace/active/{task-id}/task.md`：
@@ -109,7 +109,7 @@ node .agents/scripts/validate-artifact.js gate watch-pr .agents/workspace/active
   ```
   下一步 - 完成并归档任务：
     - Claude Code / OpenCode：/complete-task {task-ref}
-    - Gemini CLI：/agent-infra:complete-task {task-ref}
+    - Gemini CLI：/{{project}}:complete-task {task-ref}
     - Codex CLI：$complete-task {task-ref}
   ```
 

package/templates/.agents/skills/watch-pr/reference/monitor-and-heal.en.md

@@ -12,6 +12,13 @@ After running the watch command from `.agents/rules/pr-checks-commands.md`, clas
 
 ## Self-Heal Decision Tree
 
+```text
+# self-heal-test-command-contract
+primary: failing-job-command
+fallback-source: project-test-skill
+unknown: help
+```
+
 For each failing check, decide "self-heal" vs "ask for help" in this order:
 
 1. **Can the corresponding CI run be located** (per the rule's "Resolve a failing run id")? No → ask for help.
@@ -22,7 +29,7 @@ For each failing check, decide "self-heal" vs "ask for help" in this order:
 4. When "locatable + code layer + under cap" holds, perform one self-heal:
    - Before fixing, run `git status -s` to record the working tree and ensure only changes related to this failure are included.
    - Locate and make a minimal fix per the logs (touch only code / tests / config related to that failure).
-   - Run the relevant tests: prefer the local command for the failing job; fall back to `npm run test:core` repo-wide. **Do not commit or push before tests pass.**
+   - Run the relevant tests: prefer the local command for the failing job; otherwise read the project `test` skill and select its declared core or full validation command. If neither command is known, ask for help. **Do not commit or push before tests pass.**
    - After tests pass, publish the fix: per `.agents/rules/commit-and-pr.md` stage only the related files (`git add <paths>`, avoid `git add -A` pulling in unrelated changes) → create the fix commit (`git commit`) → `git push` to the current PR branch.
    - Record the fix commit SHA, increment the fix count, and return to SKILL step 2 to re-watch.
    - Never make unrelated "drive-by" optimizations; never loosen / skip the failing assertion to "make it green".

package/templates/.agents/skills/watch-pr/reference/monitor-and-heal.zh-CN.md

@@ -12,6 +12,13 @@
 
 ## 自愈决策树
 
+```text
+# self-heal-test-command-contract
+primary: failing-job-command
+fallback-source: project-test-skill
+unknown: help
+```
+
 对每个失败 check，按下列顺序判定「自愈」还是「求助」：
 
 1. **能否定位到对应的 CI run**（按规则的「解析失败 run id」）？否 → 求助。
@@ -22,7 +29,7 @@
 4. 满足「可定位 + 代码层 + 未达上限」时执行一次自愈：
    - 自愈前先 `git status -s` 记录当前工作树，确保后续只纳入与本次失败相关的改动。
    - 在本地按日志定位并最小化修复（只动与该失败相关的代码 / 测试 / 配置）。
-   - 运行对应测试：优先失败 job 对应的本地命令；仓库统一回退 `npm run test:core`。**测试通过前不得提交或推送。**
+   - 运行对应测试：优先失败 job 对应的本地命令；否则读取项目 `test` skill，选择其中声明的 core 或 full 验证命令。两者均未知时进入求助出口。**测试通过前不得提交或推送。**
    - 测试通过后发布修复：按 `.agents/rules/commit-and-pr.md` 仅暂存本次相关文件（`git add <相关路径>`，避免 `git add -A` 卷入无关改动）→ 创建修复 commit（`git commit`）→ `git push` 到当前 PR 分支。
    - 记录本次修复的 commit SHA，修复计数 +1，回到 SKILL 步骤 2 重新监控。
    - 绝不执行与失败无关的「顺手优化」；不放宽 / 跳过失败的断言来「修绿」。