@elvis1513/auto-coding-skill 4.1.9 → 4.2.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -9,6 +9,13 @@ The Skill is a selectable guardrail, not a command sequence that must run for
9
9
  every task. The model skips machinery whose expected benefit does not exceed its
10
10
  cost; read-only work and obvious small clean-checkout changes normally stay direct.
11
11
 
12
+ Version 4.2.1 makes dirty/concurrent classification fail closed, bounds focused
13
+ review latency, provides a complete Reviewer result template, preserves the
14
+ caller's runtime environment for routed commands, keeps non-authoritative Agent
15
+ history out of structure findings, and distinguishes UI paths and mechanical
16
+ renames from semantic changes. Version 4.2.0 makes registered closure resilient
17
+ to normal Git staging, reuses an exact successful final gate, and lets an active
18
+ task expand scope safely before its first commit.
12
19
  Version 4.1.9 makes `autocoding init` verify the bounded managed installation after every install or upgrade:
13
20
  it replaces every managed constraint, migrates only schema-approved project
14
21
  configuration, and converges `docs/` to one canonical directory framework. Version 4.1.7 made
@@ -74,7 +81,12 @@ The result includes:
74
81
  - `execution_mode=direct`: clean checkout and one writer; stay on the current branch.
75
82
  - `execution_mode=isolated`: dirty checkout, explicit isolation, or multiple writers.
76
83
  - `execution_mode=none`: no task signal or change; create no branch/worktree.
84
+ - `repo`, `workspace_dirty`, `dirty_paths`, and `active_writer`: the repository
85
+ and concurrency evidence used for the execution decision.
77
86
  - `review_required` and `design_required`: risk-based escalation decisions.
87
+ - `change_nature`: `mechanical`, `semantic`, or conservative `unknown`; a
88
+ mechanical rename may skip a generic durable-design requirement but never a
89
+ project rule or required high-risk review.
78
90
  - `task_kind`: `read_only`, `change`, `terminal_maintenance`, or internal `none`.
79
91
  - `mechanism_plan.required`: the complete minimum mechanism set for the task.
80
92
  - `mechanism_plan.optional_when_beneficial`: model-selectable mechanisms whose
@@ -84,18 +96,27 @@ The result includes:
84
96
  never automatic stages.
85
97
  - an adaptive agent plan that does not force fan-out for ordinary work.
86
98
 
99
+ Classification reads the working tree with one fail-closed `git status
100
+ --porcelain=v2 -z --untracked-files=all --no-renames` snapshot. A Git error or
101
+ malformed result blocks classification instead of being treated as clean. Both
102
+ `direct` and dirty/isolated decisions are point-in-time verdicts: reclassify if
103
+ the checkout or writer state changes before the first write. Any other active
104
+ direct claim is a writer, including another claim with the same owner identity.
105
+
87
106
  Clean serial work proceeds directly on the current branch with normal Git; it does
88
107
  not create machine lifecycle state. Use `task-start` only when classification
89
108
  requires isolation/review or the user explicitly requests lifecycle tracking:
90
109
 
91
- When a direct task materially changes scope after writing, reclassify with
92
- `--continue-direct` and repeat every current task `--planned-path`. Undeclared
93
- dirty paths, another writer, or mandatory isolation still selects a worktree. If
94
- the new plan requires review lifecycle, reuse `--continue-direct` on `task-start`.
110
+ If a registered task discovers another path before its first commit, expand it
111
+ with `task-scope-add` instead of finishing and restarting. Expansion is explicit,
112
+ lease/conflict checked, risk-monotonic, and invalidates old review/gate state.
113
+ Unregistered direct work should reclassify before writing outside its clean claim.
95
114
 
96
115
  ```bash
97
116
  python3 docs/tools/autopipeline/ap.py task-start T0001 \
98
117
  --owned-path backend/internal/orders --force-lifecycle
118
+ python3 docs/tools/autopipeline/ap.py task-scope-add T0001 \
119
+ --owned-path config/orders
99
120
  python3 docs/tools/autopipeline/ap.py commit-push T0001 \
100
121
  --msg "T0001: fix order retry"
101
122
  ```
@@ -163,6 +184,10 @@ commands fail before staging. Documentation-only changes may use the built-in
163
184
  diff check. `git diff --check` remains an additional hygiene check and is never
164
185
  treated as business validation.
165
186
 
187
+ Route commands run in a non-login Bash that inherits the invoking process PATH
188
+ and selected runtime. Projects that intentionally require profile activation
189
+ declare it inside the configured command.
190
+
166
191
  Check coverage without running project commands:
167
192
 
168
193
  ```bash
@@ -175,6 +200,10 @@ closure gate is limited to one stable-diff run. Full regression, Docker, builds,
175
200
  Jenkins, deployment, API verification, and target checks remain explicit
176
201
  diagnostics outside normal closure.
177
202
 
203
+ Registered tasks call `commit-push` directly because it performs or strictly
204
+ reuses that final gate. Run `light-gate` manually only for unregistered normal Git
205
+ closure or when explicitly requesting a fresh diagnostic pass.
206
+
178
207
  The recommended final closure defaults are 120 seconds per command and 180
179
208
  seconds total. Projects may raise them for measured affected-scope checks or set
180
209
  a smaller `timeout_seconds` on a route. A timeout calls for a narrower route and
@@ -188,10 +217,29 @@ never triggers a broader fallback.
188
217
  - Parallel fixers require separate worktrees and non-overlapping paths.
189
218
  - Reviewer fingerprint approval is required for high-risk, cross-module,
190
219
  explicitly configured, or parallel integration work.
220
+ - The `reviewer` Agent remains `xhigh`. A focused review gets one 90-second pass
221
+ over the supplied stable diff and relevant evidence. Timeout produces
222
+ `blocked`; do not call `task-review` for that result.
223
+ - Cross-module or parallel work and API, auth, database, payment, file-transfer,
224
+ gateway, or production-configuration boundaries use 300-second deep review.
225
+ Fixing Reviewer JSON formatting reuses the same analysis and never starts
226
+ another substantive review.
191
227
  - DD/ADR is created only for lasting cross-module, API, data, security,
192
228
  deployment, or key user-flow decisions.
193
229
  - Historical debt does not block product work unless the current change worsens it.
194
230
 
231
+ Before returning a Reviewer result, generate the complete 16-field skeleton from
232
+ the validated assignment, then fill its summary, evidence, findings, and risks:
233
+
234
+ ```bash
235
+ python3 docs/tools/autopipeline/ap.py agent-result-template \
236
+ --file <assignment.json> --verdict <approved|changes-requested|blocked>
237
+ ```
238
+
239
+ The generated object prevents field-by-field contract retries; contract checking
240
+ also reports all known field errors together. If only its JSON shape needs repair,
241
+ correct that object without rerunning review analysis.
242
+
195
243
  The main agent owns architecture, final validation, Git closure, ordered
196
244
  integration, push, and cleanup. Push ends the coding task; later CI/acceptance is
197
245
  not polled automatically. An explicitly requested failure diagnosis may continue
@@ -214,6 +262,43 @@ schema/body, runtime launcher, and documentation framework. It preserves explici
214
262
  model overrides and current values at supported project/access/concurrency/route
215
263
  fields. Removed content is archived outside active docs.
216
264
 
265
+ ## What changed in 4.2.1
266
+
267
+ - Made classification take one fail-closed Git porcelain-v2 snapshot and expose
268
+ `repo`, `workspace_dirty`, `dirty_paths`, and `active_writer` in normal output.
269
+ Direct/dirty decisions are explicit snapshots, and another same-owner direct
270
+ claim still counts as a concurrent writer.
271
+ - Kept the `reviewer` Agent at `xhigh` while adding one-pass, 90-second focused
272
+ review and 300-second deep routing for cross-module, parallel, API, auth,
273
+ database, payment, file-transfer, gateway, and production-configuration changes.
274
+ Focused timeout is `blocked`, never an approval or a reason to call `task-review`.
275
+ - Added `agent-result-template` to generate all 16 Reviewer result fields from a
276
+ validated assignment and made contract validation return all known field errors
277
+ together. JSON-only repair reuses the completed analysis instead of launching
278
+ another substantive review.
279
+ - Replaced the login Shell used for routed commands with an inherited-environment
280
+ non-login Shell, preserving the caller's PATH and selected Python/Node runtime.
281
+ - Excluded `.agents/archive/**`, managed Agent definitions, and the install
282
+ manifest from structure scans while continuing to scan project-owned `.agents` code.
283
+ - Replaced substring UI classification with exact directory segments and UI-only
284
+ extensions, removing false browser signals from files such as `.env.components`.
285
+ - Added conservative mechanical/semantic/unknown change classification. Exact
286
+ renames and explicitly behavior-stable syncs can avoid generic cross-module
287
+ design ceremony; semantic/unknown changes and explicit project design rules
288
+ remain protected, and high-risk review is unchanged.
289
+
290
+ ## What changed in 4.2.0
291
+
292
+ - Made `commit-push` preserve already staged deletions and mixed index/worktree
293
+ state, and made ownership/validation see both ends of a rename.
294
+ - Added an exact per-worktree final-gate PASS receipt. Registered `commit-push`
295
+ reuses it only when content, index, base, scope, validation plan, command hashes,
296
+ and writer lease still match; staging or push retries no longer rerun a stable gate.
297
+ - Added `task-scope-add` for conflict-checked, monotonic scope expansion before the
298
+ first commit. Expansion raises risk when necessary and invalidates review/gate state.
299
+ - Clarified that registered tasks call `commit-push` directly; standalone
300
+ `light-gate` is for unregistered Git closure or an explicit fresh diagnostic.
301
+
217
302
  ## What changed in 4.1.9
218
303
 
219
304
  - Added a packaged managed-install manifest with path, version, content hash, and
@@ -8,21 +8,43 @@ developer_instructions = """
8
8
 
9
9
  工作要求:
10
10
  1. 只审查主代理给定的 task_id、base_sha、diff_base、diff_head、diff_fingerprint 和 owning_fixer;确认 diff 稳定后再开始。
11
- 2. 优先检查:正确性、安全性、边界条件、异常路径、并发风险、事务一致性、缓存一致性、兼容性。
12
- 3. 结构/优化 review 仅在已有 baseline/backlog 时读取;可选文档缺失不构成 blocker,不重复报告 accepted debt。
13
- 4. 检查实际 diff 和已执行验证是否一致;普通任务不要求 taskbook、closure 或 evidence 文档。
14
- 5. 涉及鉴权、权限、支付、文件上传下载、部署、依赖或数据边界时,优先使用可用的安全扫描 / security review 能力辅助复核。
15
- 6. 以 git diff、相关测试、接口契约、真实运行证据、结构标准为基础,不凭风格偏好输出问题。
16
- 7. 优先输出真实问题,不要输出样式洁癖式意见。
17
- 8. 每条问题尽量包含:
11
+ 2. 若主代理指定 focused review,只对给定的稳定 diff 和相关 evidence 做一次实质分析,不扩展为全库扫描。90 秒超时即返回 `status=blocked`、`verdict=blocked`;此时不得调用 `task-review`。
12
+ 3. 跨模块、并行 writer,以及 API、鉴权、数据库、支付、文件传输、网关、生产配置边界使用 deep review,预算 300 秒;仍然只审查主代理交付的范围。
13
+ 4. 优先检查:正确性、安全性、边界条件、异常路径、并发风险、事务一致性、缓存一致性、兼容性。
14
+ 5. 结构/优化 review 仅在已有 baseline/backlog 时读取;可选文档缺失不构成 blocker,不重复报告 accepted debt。
15
+ 6. 检查实际 diff 和已执行验证是否一致;普通任务不要求 taskbook、closure 或 evidence 文档。
16
+ 7. 涉及鉴权、权限、支付、文件上传下载、部署、依赖或数据边界时,优先使用可用的安全扫描 / security review 能力辅助复核。
17
+ 8. 以 git diff、相关测试、接口契约、真实运行证据、结构标准为基础,不凭风格偏好输出问题。
18
+ 9. 优先输出真实问题,不要输出样式洁癖式意见。
19
+ 10. 每条问题尽量包含:
18
20
  - 问题点
19
21
  - 影响范围
20
22
  - 复现条件
21
23
  - 建议修法
22
- 9. 发现测试不足时,指出应补哪些测试或验证路径。
23
- 10. 结论必须是 approved 或 changes-requested;changes-requested 只返回 owning_fixer,修改后必须绑定新的 diff fingerprint。代码、配置或行为语义变化需要重新审查;仅链接、归档索引、状态措辞等机械文档修正,可由同一 reviewer 只做 targeted recheck,不必重跑完整审计。
24
- 11. 不修改代码,不提交、不推送、不集成、不清理分支;审查不扩展本地门禁。
25
- 12. 最终只返回一个 JSON 对象,不要 Markdown 或代码围栏。对象必须符合 `.agents/skills/auto-coding-skill/data/contracts/orchestration-v1.schema.json#/$defs/agentResult`,包含全部必填字段,`contract_version` 固定为 1,`role` 固定为 `reviewer`,`changed_paths` 固定为空数组,`diff_fingerprint` 必须原样返回;审查完成时 `status=completed` 且 `verdict` 只能是 `approved` 或 `changes-requested`,无法继续时使用 `status=blocked`、`verdict=blocked`。findings 按 P0/P1/P2/P3 排序。
26
- 13. 只阻断当前任务承诺范围内由本次变更新增或恶化的问题。相邻或既有问题写入 risks/follow-up,不得无限扩大当前任务和 review 范围。
27
- 14. 目标分支推送成功后,不自动跟踪 Jenkins、部署或项目负责人的实际验收;用户明确要求诊断本次推送引发的失败时,可在同一会话和任务范围继续,不要求人为新建台账任务。
24
+ 11. 发现测试不足时,指出应补哪些测试或验证路径。
25
+ 12. 实质结论必须是 approved、changes-requested 或 blocked;changes-requested 只返回 owning_fixer,修改后必须绑定新的 diff fingerprint。代码、配置或行为语义变化需要重新审查;仅链接、归档索引、状态措辞等机械文档修正,可由同一 reviewer 只做 targeted recheck,不必重跑完整审计。
26
+ 13. 不修改代码,不提交、不推送、不集成、不清理分支;审查不扩展本地门禁。
27
+ 14. 输出前运行 `agent-result-template --file <assignment.json> --verdict <approved|changes-requested|blocked>`,从已验证 assignment 生成包含全部 16 个必填字段的骨架,然后填入本次分析的 summary、evidence、findings 和 risks。最终只返回该 JSON 对象,不要 Markdown 或代码围栏。最小完整形状为:
28
+ {
29
+ "contract_version": 1,
30
+ "node_id": "<assignment.node_id>",
31
+ "role": "reviewer",
32
+ "task_id": "<assignment.task_id>",
33
+ "base_sha": "<assignment.base_sha>",
34
+ "status": "completed",
35
+ "summary": "<review summary>",
36
+ "depends_on": [],
37
+ "owned_paths": [],
38
+ "changed_paths": [],
39
+ "diff_fingerprint": "<assignment.diff_fingerprint>",
40
+ "evidence": [],
41
+ "findings": [],
42
+ "verdict": "approved",
43
+ "risks": [],
44
+ "next_owner": "main"
45
+ }
46
+ `contract_version` 固定为 1,`role` 固定为 `reviewer`,`changed_paths` 固定为空数组,`diff_fingerprint` 必须原样返回;审查完成时 `status=completed`,无法继续时 `status=blocked`。`approved` 和 `blocked` 的 `next_owner` 使用 `main`,`changes-requested` 使用 owning_fixer,findings 按 P0/P1/P2/P3 排序。
47
+ 15. 若合约检查只报告 JSON 字段、类型或格式问题,只修正上述已生成对象并复用同一次分析;不重读 diff、不重跑测试、不开启第二次实质审查。
48
+ 16. 只阻断当前任务承诺范围内由本次变更新增或恶化的问题。相邻或既有问题写入 risks/follow-up,不得无限扩大当前任务和 review 范围。
49
+ 17. 目标分支推送成功后,不自动跟踪 Jenkins、部署或项目负责人的实际验收;用户明确要求诊断本次推送引发的失败时,可在同一会话和任务范围继续,不要求人为新建台账任务。
28
50
  """
@@ -1 +1 @@
1
- {"schema_version":1,"skill_version":"4.1.9","manifest_path":".agents/managed-install.json","entries":[{"path":".agents/agents/browser-debugger.toml","source":"agents/browser-debugger.toml","ownership":"normalized-agent","sha256":"40727952944917eb0b573bb96dca2906c27bc794737f3b6b392b56df7b4208aa","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/agents/docs-researcher.toml","source":"agents/docs-researcher.toml","ownership":"normalized-agent","sha256":"aa1fed6a6a1cde282fc152c7a88b8e097a6c53bcf447d35978607d594e3cb3e4","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/agents/explorer.toml","source":"agents/explorer.toml","ownership":"normalized-agent","sha256":"4cd89888acf7c0c04fbbcd0b9a0f346c6efed42fb723ec2037af9331a84a1cac","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/agents/fixer.toml","source":"agents/fixer.toml","ownership":"normalized-agent","sha256":"7bc00a45b9d1bcc90b8b0f06d773556eba1d52effff664472c3cb81073f6abfb","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/agents/reviewer.toml","source":"agents/reviewer.toml","ownership":"normalized-agent","sha256":"f3a0adc06bba44ea81e1b10c6e353f53ece555c55f9e2990276f4c12a483f318","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/SKILL.md","source":"skill/SKILL.md","ownership":"exact","sha256":"06284b059b247000d9d0f5cc973199df99451c5c7100382ce31893578c3a2e87","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/data/contracts/orchestration-v1.schema.json","source":"skill/data/contracts/orchestration-v1.schema.json","ownership":"exact","sha256":"3a93c2bfbb236e5bc41b7ab75a67ff56e7ad3535adc229518b45c2f9fd8707e3","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/data/policies/workflow-migrations-v1.json","source":"skill/data/policies/workflow-migrations-v1.json","ownership":"exact","sha256":"1180d44ddbe455e4d38626c9ce5b1fdb6612ba0869a4cb688ee9b7a851b3bf0d","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/data/templates/ENGINEERING.md","source":"skill/data/templates/ENGINEERING.md","ownership":"exact","sha256":"8a87ef56d909cb88a31e0dcd9fe6d26ba017128fe3665c2a90b4b2cead3fc944","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/data/templates/bridges/AGENTS.md","source":"skill/data/templates/bridges/AGENTS.md","ownership":"exact","sha256":"e3fa266a7bac4d0094d541028f5a2438fd3f44b29c85cd828f8dbd236d2fa0ce","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/data/templates/tools/ap.py","source":"skill/data/templates/tools/ap.py","ownership":"exact","sha256":"c352fbf2cbcb9692a34fd952d8541b1cbe6d475c6e11e64500a1f6b5283a00be","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/requirements.txt","source":"skill/requirements.txt","ownership":"exact","sha256":"0969da99a0bc2a1b71ed50584560f4588a37567ac63af3ddbaf3c4617ca5621a","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/scripts/ap.py","source":"skill/scripts/ap.py","ownership":"exact","sha256":"56f58bffe1c59b55a1ffdcff781dbad01f675a7b248a7ace8eec06bc7d6b4901","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/scripts/core.py","source":"skill/scripts/core.py","ownership":"exact","sha256":"56019d529e82767f6dce8505945c55fc2253b596200b0e915fc89787a3bc4837","executable":true,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/scripts/http_checks.py","source":"skill/scripts/http_checks.py","ownership":"exact","sha256":"5620e187a0129f135621425a84b3a4e8f31137263c7e549a36a801fc36ff54ea","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/scripts/install_integrity.py","source":"skill/scripts/install_integrity.py","ownership":"exact","sha256":"20b0424152d61b3c067feb9558dd2532ab3abb459fc71a4c7de915e95bd83696","executable":false,"scope":"shared","version":"4.1.9"},{"path":".agents/skills/auto-coding-skill/scripts/scaffold_templates.py","source":"skill/scripts/scaffold_templates.py","ownership":"exact","sha256":"c825f15591b41403b7a97877c17f66cddcbbc5b029d143a7f1252e85388608e1","executable":false,"scope":"shared","version":"4.1.9"},{"path":"AGENTS.md","source":"skill/data/templates/bridges/AGENTS.md","ownership":"exact","sha256":"e3fa266a7bac4d0094d541028f5a2438fd3f44b29c85cd828f8dbd236d2fa0ce","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/ENGINEERING.md","source":"skill/data/templates/ENGINEERING.md#managed-workflow","ownership":"managed-workflow","sha256":"d84bd6220d637f8e91ed77fb244fec45747b64cb12910a280b9984db1ac11fb7","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/architecture/adr/_TEMPLATE-ADR.md","source":"generated:scaffold_templates:docs/architecture/adr/_TEMPLATE-ADR.md","ownership":"exact","sha256":"88c89065ae0f8555412dbe096c05342152efc7becae8d8981e6442cdfee8e672","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/architecture/structure-standard.md","source":"generated:scaffold_templates:docs/architecture/structure-standard.md","ownership":"exact","sha256":"aa2e9bbf1baad0f3a3e0e97aaaefee7282740f09a976b1bb9be7fa0b529a549a","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/deployment/deploy-records/_TEMPLATE-DEPLOY-RECORD.md","source":"generated:scaffold_templates:docs/deployment/deploy-records/_TEMPLATE-DEPLOY-RECORD.md","ownership":"exact","sha256":"f2c17c0febc322b0ad50fb9d6f7e64e6c0cfb15c8a3e445a99da40f0a0dd1901","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/design/_TEMPLATE-DD.md","source":"generated:scaffold_templates:docs/design/_TEMPLATE-DD.md","ownership":"exact","sha256":"6ca022d127bfff68521f466b60bace4f1206baf0913daa855b9f751e74198db3","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/reviews/_TEMPLATE-REVIEW.md","source":"generated:scaffold_templates:docs/reviews/_TEMPLATE-REVIEW.md","ownership":"exact","sha256":"134f0d1bd3c7a317ef3cf4c2aea9f0539f3d53dccd5bcdabf6915b3629ef8802","executable":false,"scope":"project","version":"4.1.9"},{"path":"docs/tools/autopipeline/ap.py","source":"skill/data/templates/tools/ap.py","ownership":"exact","sha256":"c352fbf2cbcb9692a34fd952d8541b1cbe6d475c6e11e64500a1f6b5283a00be","executable":false,"scope":"project","version":"4.1.9"}],"managed_namespaces":[{"path":".agents/skills/auto-coding-skill","scope":"shared"},{"path":"docs/tools/autopipeline","scope":"project"}],"preserved":[".agents/archive/**",".agents/agents/* (except paths listed in entries)","docs/project/**","docs/architecture/adr/[0-9]*.md","docs/design/*.md","docs/interfaces/*.md","docs/deployment/deploy-records/*.md","docs/reviews/*.md"]}
1
+ {"schema_version":1,"skill_version":"4.2.1","manifest_path":".agents/managed-install.json","entries":[{"path":".agents/agents/browser-debugger.toml","source":"agents/browser-debugger.toml","ownership":"normalized-agent","sha256":"40727952944917eb0b573bb96dca2906c27bc794737f3b6b392b56df7b4208aa","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/agents/docs-researcher.toml","source":"agents/docs-researcher.toml","ownership":"normalized-agent","sha256":"aa1fed6a6a1cde282fc152c7a88b8e097a6c53bcf447d35978607d594e3cb3e4","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/agents/explorer.toml","source":"agents/explorer.toml","ownership":"normalized-agent","sha256":"4cd89888acf7c0c04fbbcd0b9a0f346c6efed42fb723ec2037af9331a84a1cac","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/agents/fixer.toml","source":"agents/fixer.toml","ownership":"normalized-agent","sha256":"7bc00a45b9d1bcc90b8b0f06d773556eba1d52effff664472c3cb81073f6abfb","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/agents/reviewer.toml","source":"agents/reviewer.toml","ownership":"normalized-agent","sha256":"3b840eed1c60769f8fc8bff5cac62b1ff0dd1596e33dbb14ae33961289991894","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/SKILL.md","source":"skill/SKILL.md","ownership":"exact","sha256":"078e2b8d0b4396ad7788f579c7338bc0bf9aecc717132ee90c745286aa06f128","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/data/contracts/orchestration-v1.schema.json","source":"skill/data/contracts/orchestration-v1.schema.json","ownership":"exact","sha256":"3a93c2bfbb236e5bc41b7ab75a67ff56e7ad3535adc229518b45c2f9fd8707e3","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/data/policies/workflow-migrations-v1.json","source":"skill/data/policies/workflow-migrations-v1.json","ownership":"exact","sha256":"2db18e040260ff3e7d0108f8f9a9f2c2a39ead2ca16117f5f6b2480cce9d5c16","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/data/templates/ENGINEERING.md","source":"skill/data/templates/ENGINEERING.md","ownership":"exact","sha256":"0b493750f90b93fbb86cf3813fcf42b7e23780e93eedaa2416eaa38c1a91b033","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/data/templates/bridges/AGENTS.md","source":"skill/data/templates/bridges/AGENTS.md","ownership":"exact","sha256":"ae9912b53d26cdba3b753e00a59184127fe0850d4ce324bdabfd1c7454f9dbb0","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/data/templates/tools/ap.py","source":"skill/data/templates/tools/ap.py","ownership":"exact","sha256":"c352fbf2cbcb9692a34fd952d8541b1cbe6d475c6e11e64500a1f6b5283a00be","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/requirements.txt","source":"skill/requirements.txt","ownership":"exact","sha256":"0969da99a0bc2a1b71ed50584560f4588a37567ac63af3ddbaf3c4617ca5621a","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/scripts/ap.py","source":"skill/scripts/ap.py","ownership":"exact","sha256":"cf61031e542fd2c8f83f9ed48a3dea815b68bdbb5a1edaf79fb5f39178480920","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/scripts/core.py","source":"skill/scripts/core.py","ownership":"exact","sha256":"640c37e8b274d90d6b7765d4243e054360d5a5a85b3150fbfe6449527c1cd28f","executable":true,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/scripts/http_checks.py","source":"skill/scripts/http_checks.py","ownership":"exact","sha256":"5620e187a0129f135621425a84b3a4e8f31137263c7e549a36a801fc36ff54ea","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/scripts/install_integrity.py","source":"skill/scripts/install_integrity.py","ownership":"exact","sha256":"20b0424152d61b3c067feb9558dd2532ab3abb459fc71a4c7de915e95bd83696","executable":false,"scope":"shared","version":"4.2.1"},{"path":".agents/skills/auto-coding-skill/scripts/scaffold_templates.py","source":"skill/scripts/scaffold_templates.py","ownership":"exact","sha256":"c825f15591b41403b7a97877c17f66cddcbbc5b029d143a7f1252e85388608e1","executable":false,"scope":"shared","version":"4.2.1"},{"path":"AGENTS.md","source":"skill/data/templates/bridges/AGENTS.md","ownership":"exact","sha256":"ae9912b53d26cdba3b753e00a59184127fe0850d4ce324bdabfd1c7454f9dbb0","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/ENGINEERING.md","source":"skill/data/templates/ENGINEERING.md#managed-workflow","ownership":"managed-workflow","sha256":"c8f768515a6d89145d1ce117d21b01f2e9481819ec2f8d6838a9f8788e1b1249","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/architecture/adr/_TEMPLATE-ADR.md","source":"generated:scaffold_templates:docs/architecture/adr/_TEMPLATE-ADR.md","ownership":"exact","sha256":"88c89065ae0f8555412dbe096c05342152efc7becae8d8981e6442cdfee8e672","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/architecture/structure-standard.md","source":"generated:scaffold_templates:docs/architecture/structure-standard.md","ownership":"exact","sha256":"aa2e9bbf1baad0f3a3e0e97aaaefee7282740f09a976b1bb9be7fa0b529a549a","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/deployment/deploy-records/_TEMPLATE-DEPLOY-RECORD.md","source":"generated:scaffold_templates:docs/deployment/deploy-records/_TEMPLATE-DEPLOY-RECORD.md","ownership":"exact","sha256":"f2c17c0febc322b0ad50fb9d6f7e64e6c0cfb15c8a3e445a99da40f0a0dd1901","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/design/_TEMPLATE-DD.md","source":"generated:scaffold_templates:docs/design/_TEMPLATE-DD.md","ownership":"exact","sha256":"6ca022d127bfff68521f466b60bace4f1206baf0913daa855b9f751e74198db3","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/reviews/_TEMPLATE-REVIEW.md","source":"generated:scaffold_templates:docs/reviews/_TEMPLATE-REVIEW.md","ownership":"exact","sha256":"134f0d1bd3c7a317ef3cf4c2aea9f0539f3d53dccd5bcdabf6915b3629ef8802","executable":false,"scope":"project","version":"4.2.1"},{"path":"docs/tools/autopipeline/ap.py","source":"skill/data/templates/tools/ap.py","ownership":"exact","sha256":"c352fbf2cbcb9692a34fd952d8541b1cbe6d475c6e11e64500a1f6b5283a00be","executable":false,"scope":"project","version":"4.2.1"}],"managed_namespaces":[{"path":".agents/skills/auto-coding-skill","scope":"shared"},{"path":"docs/tools/autopipeline","scope":"project"}],"preserved":[".agents/archive/**",".agents/agents/* (except paths listed in entries)","docs/project/**","docs/architecture/adr/[0-9]*.md","docs/design/*.md","docs/interfaces/*.md","docs/deployment/deploy-records/*.md","docs/reviews/*.md"]}
@@ -27,11 +27,11 @@ python3 docs/tools/autopipeline/ap.py classify --scope auto \
27
27
  [--task-kind read_only|change|terminal_maintenance] [--claim-direct]
28
28
  ```
29
29
 
30
- Run `mechanism_plan.required`. The model may select
31
- `optional_when_beneficial` only when expected value exceeds coordination cost.
32
- Do not run `forbidden` mechanisms unless the user explicitly overrides the plan.
33
- Reclassify before changing writer count. For uncertain direct work, create
34
- `--claim-direct` while clean; continuation requires its ID and exact paths.
30
+ Run `mechanism_plan.required`; select `optional_when_beneficial` only when its
31
+ value exceeds coordination cost, and keep `forbidden` off absent user override.
32
+ `classify` is a fail-closed snapshot: inspect `repo`, `workspace_dirty`,
33
+ `dirty_paths`, and `active_writer`, and reclassify if state changes before writing.
34
+ Same-owner claims conflict unless continuing the exact ID; `change_nature` never bypasses a rule or review.
35
35
 
36
36
  Use the registered lifecycle only when classification requires isolation or
37
37
  fingerprinted review, or when the user explicitly requests it:
@@ -39,6 +39,8 @@ fingerprinted review, or when the user explicitly requests it:
39
39
  ```bash
40
40
  python3 docs/tools/autopipeline/ap.py task-start T0001 \
41
41
  --owned-path src [--isolated] [--review-required] [--continue-direct --direct-claim <ID>]
42
+ # Before the first commit, expand an active task instead of restarting it:
43
+ python3 docs/tools/autopipeline/ap.py task-scope-add T0001 --owned-path config/new-scope
42
44
  # For review-required work, obtain the current fingerprint and approve it:
43
45
  python3 docs/tools/autopipeline/ap.py task-status T0001 --json
44
46
  python3 docs/tools/autopipeline/ap.py task-review T0001 \
@@ -50,20 +52,34 @@ python3 docs/tools/autopipeline/ap.py task-integrate T0001
50
52
 
51
53
  Each parallel writer uses a task ID, worktree, lease, dependency SHAs, and distinct
52
54
  `owned_paths`. Main integrates, gates, pushes, and cleans. Validate delegated JSON
53
- with `agent-contract-check`; parallel fixers require reclassification with `--writers >1`.
55
+ with `agent-contract-check`; parallel fixers require reclassification with
56
+ `--writers >1`.
57
+
58
+ Use the existing `reviewer` Agent at `xhigh` and follow the plan: focused means one
59
+ 90-second analysis; deep gets 300 seconds for parallel/cross-module or sensitive
60
+ boundaries. Timeout is `blocked` and is not sent to `task-review`. Generate all 16
61
+ fields with `agent-result-template`; contract checking aggregates errors, and shape
62
+ repair must reuse the same analysis.
54
63
 
55
64
  ## Close with the bounded routed gate
56
65
 
66
+ For unregistered clean serial work, run the final gate and then use normal Git:
67
+
57
68
  ```bash
58
69
  python3 docs/tools/autopipeline/ap.py validation-map-check --path <PATH>
59
70
  python3 docs/tools/autopipeline/ap.py light-gate --scope changed --explain
60
71
  ```
61
72
 
73
+ For a registered task, call only `commit-push`: it owns the final gate and reuses
74
+ an exact Git-local PASS only while content, base, scope, routes, commands, and
75
+ lease match. Matching manual passes are reusable; changed state is not.
76
+
62
77
  Every changed code/config path must map to a real project command. The runtime
63
78
  de-duplicates commands and enforces route, command, and total budgets. Defaults
64
79
  are 120 seconds per command and 180 total; projects may raise them for a measured
65
80
  affected-scope check. Timeouts should narrow the route, never trigger a full build.
66
- Focused test/fix/retest loops remain allowed before the final gate.
81
+ Focused test/fix/retest loops remain allowed before the final gate. Commands inherit
82
+ the caller's PATH in non-login Bash; declare activation in the configured command.
67
83
 
68
84
  Run full regression, Docker, Jenkins, deployment, browser/device writes, or target
69
85
  acceptance only as an explicit diagnostic. A requested just-pushed failure
@@ -71,20 +87,17 @@ diagnosis continues in the same task without inventing another lifecycle.
71
87
 
72
88
  ## Keep state and upgrades safe
73
89
 
74
- Machine task state, leases, fingerprints, and gate evidence stay in Git local
75
- state. Ordinary work creates no taskbook, closure, evidence JSONL, active-task,
76
- or design artifact. Terminal ledger reconciliation validates and commits once.
90
+ Machine task state, leases, fingerprints, and gate evidence stay Git-local.
91
+ Ordinary work creates no taskbook, closure, evidence JSONL, active-task, or
92
+ design artifact. Terminal ledger reconciliation validates and commits once.
77
93
 
78
94
  ```bash
79
95
  autocoding init
80
96
  python3 docs/tools/autopipeline/ap.py doctor
81
97
  ```
82
98
 
83
- Finish registered tasks before changing versions. `autocoding init` is idempotent
84
- and performs the complete install/upgrade: it replaces every managed constraint,
85
- preserves only schema-supported project values, installs the canonical docs
86
- directory framework and managed templates, and archives obsolete content outside
87
- active `docs/` while preserving valid architecture, ADR, interface, DD, review,
88
- and deploy-record artifacts in their designated directories. It then verifies the
89
- packaged managed-install manifest; `doctor` cheaply rechecks the same declared
90
- paths without treating project-owned files or archives as a release mirror.
99
+ Finish registered tasks before changing versions. `autocoding init` idempotently
100
+ replaces managed constraints, preserves schema-supported values, installs canonical
101
+ docs/templates, archives obsolete content outside active `docs/`, and preserves
102
+ designated artifacts. It verifies the managed-install manifest; `doctor` cheaply
103
+ rechecks declared paths without treating project-owned files or archives as a mirror.
@@ -1,8 +1,8 @@
1
1
  {
2
2
  "schema_version": 1,
3
3
  "managed_versions": {
4
- "agents": "4.1.9",
5
- "engineering": "4.1.9"
4
+ "agents": "4.2.1",
5
+ "engineering": "4.2.1"
6
6
  },
7
7
  "known_official_agents_sha256": [
8
8
  "ebaaadab60327b3a4852948443d7191a4d48cb00bf4910038ab78c9149089373"
@@ -1,15 +1,13 @@
1
1
  ---
2
2
  workflow:
3
- skill_version: "4.1.9"
3
+ skill_version: "4.2.1"
4
4
  mode: "dev"
5
5
  profile: "auto"
6
6
  completion: "push"
7
-
8
7
  project:
9
8
  name: ""
10
9
  repo_root: "."
11
10
  stack: "generic"
12
-
13
11
  access:
14
12
  project:
15
13
  frontend:
@@ -38,7 +36,6 @@ access:
38
36
  url: ""
39
37
  username: ""
40
38
  password: ""
41
-
42
39
  concurrency:
43
40
  isolation: "adaptive"
44
41
  base_ref: ""
@@ -48,12 +45,10 @@ concurrency:
48
45
  cleanup_merged: true
49
46
  delete_remote_branch: true
50
47
  disposable_ignored: []
51
-
52
48
  commands:
53
49
  # Optional convenience command. Define the project-native fast commands used
54
50
  # by validation.routes; sync fills this only for package.json test:changed.
55
51
  project_fast: ""
56
-
57
52
  validation:
58
53
  on_unmapped: "error"
59
54
  # Recommended fast defaults. Raise only for a measured affected-scope check;
@@ -63,10 +58,8 @@ validation:
63
58
  # Initialization must replace this with real path-to-command mappings. Empty
64
59
  # routes intentionally fail closed for code/config changes.
65
60
  routes: []
66
-
67
61
  risk:
68
62
  rules: []
69
-
70
63
  structure:
71
64
  enabled: true
72
65
  enforcement: "advisory"
@@ -76,12 +69,10 @@ structure:
76
69
  enabled: true
77
70
  block: false
78
71
  rules: []
79
-
80
72
  optimization:
81
73
  completion_policy: "no-new-debt"
82
74
  require_baseline_for_global_review: false
83
75
  report_accepted_debt_as_findings: false
84
-
85
76
  docs:
86
77
  framework: "engineering-centered"
87
78
  design_dir: "docs/design"
@@ -90,38 +81,29 @@ docs:
90
81
  api_doc: "docs/interfaces/api.md"
91
82
  api_change_log: "docs/interfaces/api-change-log.md"
92
83
  ---
93
-
94
84
  # Engineering Configuration and Project Facts
95
-
96
- <!-- auto-coding-skill:managed-workflow:start version=4.1.9 -->
97
-
85
+ <!-- auto-coding-skill:managed-workflow:start version=4.2.1 -->
98
86
  Root `AGENTS.md` is the single shared behavioral protocol. The installed
99
87
  `SKILL.md` contains invocation guidance. This file is the exact-schema source for
100
88
  project workflow configuration and access values that neither file can infer.
101
-
102
89
  The frontmatter contract is:
103
-
104
90
  - `workflow`: installed version and push-completion mode.
105
91
  - `access`: required project, Jenkins, GitLab, and Nexus access values.
106
92
  - `concurrency`: target branch and adaptive isolation settings.
107
93
  - `commands`: project-native executable commands.
108
94
  - `validation`: explicit path-to-command routes plus the bounded final-gate time
109
95
  budget. Code/config paths must be mapped; route-level `timeout_seconds` can
110
- bound a slower command without changing the project defaults.
96
+ bound a slower command; commands inherit PATH and declare intentional activation.
111
97
  - `risk`: project-specific signals that increase design/review depth but never
112
98
  expand the automatic local gate.
113
99
  - `structure` and `optimization`: advisory architecture and no-new-debt policy.
114
100
  - `docs`: locations inside the managed documentation directory framework.
115
-
116
101
  Run `python3 docs/tools/autopipeline/ap.py doctor` after changing frontmatter; it
117
102
  also performs the bounded local managed-install integrity check.
118
103
  Do not put competing delivery flows, agent roles, lifecycle rules, gate semantics,
119
104
  or completion rules outside this managed block.
120
-
121
105
  <!-- auto-coding-skill:managed-workflow:end -->
122
-
123
106
  # Project Facts Location
124
-
125
107
  Keep durable project facts in the installed `docs/project/` files. Do not append
126
108
  workflow rules or additional sections here; initialization owns this document's
127
109
  schema and body.
@@ -1,4 +1,4 @@
1
- <!-- auto-coding-skill:managed-agents:start version=4.1.9 -->
1
+ <!-- auto-coding-skill:managed-agents:start version=4.2.1 -->
2
2
  # Shared Engineering Protocol
3
3
 
4
4
  This file is fully managed by `auto-coding-skill`. Keep project-specific facts,
@@ -34,17 +34,15 @@ The default budget is:
34
34
  | High-risk or contract-crossing change | necessary design and independent review |
35
35
  | Multiple independent writers | one task/worktree/lease per writer, ordered integration |
36
36
 
37
- Do not add classify, task lifecycle, durable design, subagents, reviewer, or a
38
- worktree merely because the tool exists. If task kind or impact is unclear, run
39
- `classify`: run `mechanism_plan.required`, let the model select
40
- `optional_when_beneficial` only when expected value exceeds coordination cost,
41
- and keep `forbidden` mechanisms off unless the user explicitly overrides the
42
- plan. Reclassify only after a material task-kind, scope, risk, or writer change.
37
+ Do not add classify, lifecycle, design, subagents, review, or a worktree merely
38
+ because it exists. If task kind or impact is unclear, run `classify`, obey required
39
+ and forbidden mechanisms, and select optional ones by expected value. Classification
40
+ uses one fail-closed `git status --porcelain=v2 -z --untracked-files=all --no-renames`
41
+ snapshot; inspect `repo`, `workspace_dirty`, `dirty_paths`, and `active_writer`.
42
+ Git errors/malformed status block, and state changes before writing require reclassification.
43
43
 
44
- Normal delivery is:
45
-
46
- `analysis → decomposition → necessary design → development → one bounded final
47
- changed-scope gate → commit/push`.
44
+ Normal delivery is `analysis → decomposition → necessary design → development →
45
+ one bounded final changed-scope gate → commit/push`.
48
46
 
49
47
  Push to the target branch ends normal coding. Do not poll Jenkins, deploy, or run
50
48
  owner acceptance automatically. A user-requested diagnosis of the just-pushed
@@ -52,12 +50,16 @@ failure may continue in the same conversation/task without a second lifecycle.
52
50
 
53
51
  ## Git and parallel work
54
52
 
55
- - One writer in a clean checkout works directly. If review may be needed, create a
56
- claim before the first write; continuation requires that claim and exact paths.
53
+ - One writer in a clean checkout works directly. Claim uncertain direct work before
54
+ writing; any other claim, including the same owner, conflicts unless continuing its exact ID.
57
55
  - Existing unrelated changes, another writer, or configured mandatory isolation
58
56
  requires a registered task branch/worktree.
59
57
  - Every delegated fixer or parallel writer owns a task ID/worktree, writer lease,
60
58
  non-overlapping `owned_paths`, and prerequisite commit SHAs.
59
+ - Before its first commit, an active task may add explicit paths with
60
+ `task-scope-add`; the command checks leases/conflicts, only raises risk, and
61
+ invalidates prior review and final-gate state. Never restart a no-diff task just
62
+ to enlarge its declared scope.
61
63
  - Never let multiple writing agents share a checkout. Never restore, reset, stash,
62
64
  clean, overwrite, or commit unknown changes.
63
65
  - No diff means no commit, push, or temporary branch.
@@ -77,6 +79,11 @@ failure may continue in the same conversation/task without a second lifecycle.
77
79
  - Validate delegated assignment/result JSON with `agent-contract-check` before use.
78
80
  - Require fingerprinted review only for path/rule-confirmed high-risk,
79
81
  contract-crossing, parallel, or configured work; intent words are candidates only.
82
+ - Keep `reviewer` at `xhigh`: focused is one 90-second analysis; deep gets 300
83
+ seconds for parallel/cross-module or sensitive boundaries. Timeout is `blocked`
84
+ and is not sent to `task-review`.
85
+ - Generate all 16 fields with `agent-result-template`; contract checking aggregates
86
+ errors, and JSON shape repair reuses the same analysis.
80
87
  - Review blocks only defects introduced or worsened in the promised scope.
81
88
  Adjacent existing issues are non-blocking follow-ups.
82
89
  - Semantic changes invalidate approval. Mechanical documentation-only corrections
@@ -95,6 +102,10 @@ failure may continue in the same conversation/task without a second lifecycle.
95
102
  validation.
96
103
  - Focused tests may run and rerun during development. Run one final routed gate
97
104
  after the diff is stable.
105
+ - For registered tasks, `commit-push` owns that final gate; do not run a separate
106
+ `light-gate` first. A strict Git-local PASS receipt may be reused only while the
107
+ content/index, base, scope, route/command plan, and writer lease still match.
108
+ Unregistered clean serial work runs `light-gate` before normal Git closure.
98
109
  - The final gate defaults to 120 seconds per command and 180 seconds total. A
99
110
  project may raise either budget for a measured affected-scope check, and may set
100
111
  a smaller `timeout_seconds` on an individual route. A timeout should narrow the