@tea-agent/loop-agent 0.14.0 → 0.15.0

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.
Files changed (37) hide show
  1. package/CHANGELOG.md +48 -9
  2. package/README.md +23 -4
  3. package/dist/workflows/dag/init-hybrid.js +27 -11
  4. package/docs/README.md +1 -0
  5. package/docs/init-surface.manifest.json +4 -7
  6. package/docs/templates/branch-merge-report.md +0 -1
  7. package/harness.json +1 -1
  8. package/package.json +6 -16
  9. package/skills/browser-tools/SKILL.md +2 -2
  10. package/skills/frontend-design-review/references/review-checklist.md +27 -45
  11. package/skills/frontend-implementation/references/node-contracts.md +4 -4
  12. package/skills/frontend-review/SKILL.md +3 -1
  13. package/skills/frontend-review/references/review-findings.md +2 -1
  14. package/skills/frontend-verification/SKILL.md +3 -1
  15. package/skills/frontend-verification/references/verification-checklist.md +13 -22
  16. package/skills/loop-agent/references/hybrid-dag.md +1 -1
  17. package/docs/agent-dag-recovery-playbook.md +0 -195
  18. package/docs/agent-dag-runner.md +0 -67
  19. package/docs/cursor-prompt-sidecar.md +0 -36
  20. package/docs/decisions/README.md +0 -18
  21. package/docs/design/README.md +0 -167
  22. package/docs/development-principles.md +0 -73
  23. package/docs/exec-plans/README.md +0 -6
  24. package/docs/exec-plans/active/README.md +0 -13
  25. package/docs/exec-plans/completed/README.md +0 -108
  26. package/docs/feature-workflow.md +0 -414
  27. package/docs/loop-agent-harness.md +0 -142
  28. package/docs/production-readiness.md +0 -96
  29. package/docs/progress/README.md +0 -81
  30. package/docs/reports/README.md +0 -163
  31. package/docs/verification-matrix.md +0 -70
  32. package/scripts/check-product-line-docs.sh +0 -29
  33. package/scripts/check-task-pool-root.sh +0 -32
  34. package/scripts/kb-graph-incremental-prepare.sh +0 -5
  35. package/scripts/kb-graph-materialize.sh +0 -4
  36. package/scripts/kb-graph-promote.sh +0 -4
  37. package/scripts/kb-query.sh +0 -5
@@ -1,142 +0,0 @@
1
- # loop-agent Harness
2
-
3
- loop-agent 提供结构化 agent 工作的本地 harness。
4
-
5
- ## Runtime 区域
6
-
7
- - `.harness/tasks/` — task source、约束、state、loop facts
8
- - `.harness/dag-runs/` — DAG run state 与 artifacts
9
- - `.harness/runs/` — one-shot tool run 证据
10
- - `.harness/cache/` — 本地 runtime 缓存
11
- - `.harness/live/` — 瞬态 live-session 文件
12
- - `.harness/task-pool/` — `agent-worker` 的 Task Pool state、batch artifacts、failure handoffs 与 Observe events;它是 `.harness/` 内的独立 Worker runtime root,默认不提交
13
- - `.harness/task-pool/states/<featureId>/<taskId>.json` — Task Pool schema v2 的 canonical state(ADR 0004);复合身份 `TaskPoolTaskRef = { featureId, taskId }`。扁平 `states/<taskId>.json` 为 legacy,不得静默解释
14
- - `.harness/dag-runs/<lifecycle>/<run-id>/contracts/backend-test-analysis.json` — `backend-test` 的 run-owned 结构化分析契约;由 deterministic schema gate 从只读 analyze 输出生成,节点记录 path/hash/schema,随 lifecycle 目录迁移且不进入 Git writeSet
15
- - `.harness/dag-runs/<lifecycle>/<run-id>/.runtime/skill-snapshot.json` — 该 run 实际使用的 resolved skill profile snapshot;由 state 中相对 ref 与原始 bytes SHA-256 锚定,随 lifecycle 目录整体迁移
16
-
17
- ## Skill 指令
18
-
19
- - `.agents/skills/loop-agent/` — loop-agent 仓库自有的 skill 指令与参考资料
20
- - `.agents/skills/agent-worker/` — Feature Packet、TaskSpec、Task Pool、controller pinning、自举 candidate canary 与 Worker recovery 的可选 outer-loop operator skill;不是 DAG node skill
21
- - `.agents/skills/<skill-name>/` — 目标项目 repo-local skill 副本;`.agents/skills/<skill-name>/` 仍是 loop-agent 源仓库和 npm 包内置回退路径
22
- - 历史 `skill/` 不再是必需仓库路径;runtime 查找仅为 `loop-agent` 保留兼容 fallback
23
-
24
- 目标项目若在 `.agents/skills/` 中提供 `SDD-requirement-analysis`、`SDD-design-analysis`、`SDD-implementation-test-review`,hybrid DAG 生成器会按节点追加对应 skill。该增强只复用 repo-local 知识与方法:loop-agent 仍拥有流程与状态控制权,embedded-mode 约束禁止 SDD 状态推进、确认门禁和归档;初始化/扫描类 SDD skill 不会自动执行。未命中时不会改变默认 DAG。命中的节点 skill 继续由同一 resolved instruction 和 run-owned snapshot 机制冻结,没有第二套 loader 或旁路。
25
-
26
- ## Package 资产
27
-
28
- npm 包包含运行与说明 harness 所需的静态能力资产:`bin/`、`dist/`、`.agents/skills/`(包括 `loop-agent` 与 `agent-worker`)、顶层 `ai_workspace/loop-agent/*.md`、`ai_workspace/loop-agent/templates/`、`examples/`、`harness.json`、`AGENTS.md`、`README.md`、`CHANGELOG.md`。
29
-
30
- 生成或历史的 task facts 不是 package 资产。`ai_workspace/loop-agent/progress/`、`ai_workspace/loop-agent/reports/`、`ai_workspace/loop-agent/exec-plans/`、`ai_workspace/loop-agent/decisions/` 下现有文件属于目标仓库历史;包只携带这些目录的 README。
31
-
32
- 从 npm 安装 `loop-agent` 并在其他项目使用时,DAG skill 指令按配置、用户或目标项目 `.agents/skills/` 解析(若存在),再回退到包内 `.agents/skills/`。目标项目可添加本地 skills,但不必复制 loop-agent 源码历史或根 `.agents/skills/` 目录即可使用默认 DAG 工作流。
33
-
34
- `scripts/self-host-canary.mjs` 是源码仓库维护脚本,不进入 npm package `files`;安装后的 package 不把它作为用户命令 surface。
35
-
36
- ## harness.json Schema
37
-
38
- `ai_workspace/loop-agent/templates/harness.schema.json` 是面向 IDE 的 `harness.json` JSON Schema。它用于补全字段、提示枚举值、解释 executor/model routing 配置,并尽早暴露拼写错误。
39
-
40
- 运行时事实源仍是 `src/governance/manifest-types.ts` 中的 Zod `harnessManifestSchema`;CLI 读取 `harness.json` 时继续通过 `loadHarnessManifest()` 解析。JSON Schema 不替代运行时校验,只服务编辑体验。
41
-
42
- `loop-agent init --profile full --merge` 会把 schema 投影到目标项目,并在生成的 `harness.json` 写入与 `governanceRoot` 匹配的 `$schema`。旧初始化项目可通过 `loop-agent init check-update` 查看缺失项,并用 `loop-agent init update --apply-safe` 安全补齐 schema 文件和 `$schema` 引用;该安全更新只允许修改 `$schema`,不会重写用户已有 harness 配置。
43
-
44
- DAG run state 会记录 runner PID、hostname 和定期 heartbeat,并为节点记录最近活动时间。`loop-agent dag status` 和 `loop-agent dag doctor --run-id <run-id> --markdown` 会将同主机且 PID 已消失的运行识别为 `orphaned`,将进程仍在但 heartbeat 过期的运行识别为 `stale`,将 runner 存活但当前节点超过 5 分钟没有状态活动标记为 `node-quiet`;跨主机运行只标记为 `unknown-host`,不会据此判断远端进程死亡。Observe 使用同一组字段过滤活跃 DAG,并在详情中显示 lifecycle 与 liveness;`node-quiet` 仍视为活跃,只提示操作者检查 session events 和 executor logs。
45
-
46
- Observe 和 `dag status` 还会给出 `effectiveStatus`、`stateConsistent` 与 `recoveryEligibility`。主状态以 lifecycle + canonical status + liveness 的综合判定为准,原始 `state.status` 只作为诊断事实展示。遇到历史 `paused/` 目录仍记录 `status=running`、或 active runner 已明确 `orphaned` 时,先运行只读检查:
47
-
48
- ```bash
49
- loop-agent dag reconcile-run --run-id <run-id>
50
- ```
51
-
52
- 确认不应继续后,必须显式选择收口含义并填写原因:
53
-
54
- ```bash
55
- loop-agent dag reconcile-run --run-id <run-id> --action supersede --reason "implementation completed elsewhere"
56
- loop-agent dag reconcile-run --run-id <run-id> --action abandon --reason "operator cancelled obsolete work"
57
- ```
58
-
59
- 收口会保存 `reconciliation.json` 原始快照,将仍为 RUNNING 的节点标为 ERROR,并把 run 归档到 `completed/`;不会把未执行节点伪造成成功。同机 runner 仍存活、仅 `node-quiet`、heartbeat stale 但 PID 仍存在、跨主机或无法证明停止时都会拒绝收口。
60
-
61
- DAG 节点的模型由 `executorModels[executor][complexity]` 路由。新 run 会将解析后的 Pi/Cursor 模型写入节点 state;Observe 读取历史 run 时会从 `run.json` 补算缺失模型。Shell 和 Static 节点不使用模型,因此模型列保持 `-`。
62
-
63
- DAG agent 节点的输出语言由 `harness.json.workflowPolicy.dag.outputLanguage` 控制,支持 `zh-CN` 和 `en`。未配置时默认 `zh-CN`;`dag run-task` 会把配置写入生成的 DagSpec,手写 DagSpec 缺少该字段时 prompt envelope 同样默认中文。语言规则只约束说明、分析、报告、总结和文档正文,不翻译代码、命令、路径、标识符、JSON 字段、VERDICT 行及其他精确协议字面量。
64
-
65
- 新 DAG run(包括 dry-run 与 init-only)会在首次 state write 前冻结所有初始 Pi task、dynamic map child template 与 loop body template 的 resolved skill profiles。profile id 包含 ordered skills、learned-pattern mode 和两个 instruction budgets;后续节点只从 snapshot 建 prompt。snapshot/hash/binding/profile 不一致时节点进入 `ERROR` / `skill-snapshot-integrity`,executor 不调用,也不回退实时 skill。旧 run 只有在 ref 与 artifact 都不存在时,才可在 resume 前创建 `legacy-resume-backfill`,且该 mode 只证明剩余节点被固定。
66
-
67
- ## Command Surface
68
-
69
- - `new-task`、`status`、`inspect`、`doctor`
70
- - `dag run-task`、`dag validate`、`run-dag`、`dag report`、`dag resume`、`dag reconcile-run`
71
- - `plan list`、`plan create`、`plan complete`、`plan check`、`docs audit`、`docs archive`
72
- - `delegate`、`harvest`、`promote-run`、`closeout`
73
- - `loop init`、`loop run`、`loop status`、`loop closeout`
74
- - `pi-prompt`、`cursor-prompt`
75
- - `handoff check`、`spine audit`、`knowledge curate`
76
- - `agent-worker feature review|run|verify-final|delivery|closeout|approve-followup`、`agent-worker task validate|explain-profile|retry|draft-followup`、`agent-worker batch run-ready`、`agent-worker pool doctor|migrate-state`、`agent-worker report morning|metrics`、`agent-worker observe serve|snapshot`
77
-
78
- `dag run-task` 在生成 DAG 草稿前会运行与 `plan check` 同源的 exec-plan 索引校验;active plan 与 `ai_workspace/loop-agent/exec-plans/active/README.md` 索引不一致时立即失败,避免运行昂贵节点后才在末端 shell verify 发现。`plan create` 优先复用目标项目模板并回退到发布包内置 `ai_workspace/loop-agent/templates/exec-plan.md`,`plan create`/`plan complete` 的多文件写入均有回滚保护;`new-task` 不自动绑定 exec-plan。
79
-
80
- ## Feature Review
81
-
82
- `agent-worker feature review --feature-dir <feature-dir> --repo <target-repo>` 是只读的 Feature 级决策入口。它复用 Feature Packet validation、task graph 和 `.harness/task-pool/` 事实,派生 `Draft`、`Ready`、`Running`、`NeedsAction`、`AwaitingQA`、`Deliverable`、`Closed`,同时展示 required AC 覆盖、阻塞、证据和唯一主行动。
83
-
84
- 默认输出按结果、原因、下一步、证据组织;`--json` 输出 schemaVersion 1 JSON。缺失或损坏的 Packet、state、JSONL 或 closeout 会暴露 validation error / projection warning,不会被误判为无数据、可交付或已关闭。该命令不写 Feature Packet、Task Pool 或 Git。
85
-
86
- `agent-worker feature run --feature-dir <feature-dir> --repo <target-repo>` 是薄编排入口:先执行 Feature validation 和一次目标仓库 preflight,再复用 `runReadyTasks()` 推进 Ready Task,最后刷新 morning report、`.harness/task-pool/observability/snapshot.json` 与 Feature review。`--dry-run` 只读取 Packet/Task Pool 并列出 Ready IDs、步骤和预期产物。默认无 Git 授权时单次最多执行一个 Ready 写任务;checkpoint 模式可按 `--limit` 串行推进,并在每个 Task 后建立 clean boundary。
87
-
88
- 写入型 Worker 命令可传 `--expected-controller-version` 与 `--expected-controller-fingerprint`。`LoopAgentClient` 先把请求的 bin/PATH/npm shim 解析成冻结的绝对 launch spec,再校验 package name/version、binary SHA-256 和 portable package fingerprint;expected gate 在 materialize、Task Pool `Running` 或目标仓库写入前完成。预检后的命令使用该 launch spec,不会因 PATH、symlink 或安装内容变化而静默切换 controller。
89
-
90
- `agent-worker feature verify-final` 在 clean Delivery HEAD 上以已完成的 `qa-execute` TaskSpec 运行独立最终 DAG,跳过 promotion/task closeout,避免 repo write 或 HEAD 移动,并原子生成 canonical QA aggregate 与 HEAD-bound final verification。`agent-worker feature delivery` 从同一 Git transaction 和 Task Pool facts 生成带 hash 的 manifest、Acceptance Coverage 与稳定 `PR.md`,拒绝 dirty/wrong branch、历史或 trailer 漂移、缺失 run/QA/final verification 和 required AC 缺口。`agent-worker feature closeout` 默认只读预览 gates;显式 `--apply --owner` 才原子写回 Closeout,并在 stale facts、写入或 post-validation 失败时恢复原状态。
91
-
92
- Morning report 顶部和 Observe 首页直接消费共享 Feature read model,固定从 Status、Next Action、Why、Evidence 开始。`agent-worker report metrics --month YYYY-MM` 写入 `.harness/task-pool/artifacts/metrics/monthly-YYYY-MM.json|md`,每项指标保留分母、样本量、窗口和 missingData。
93
-
94
- 显式 `--git-mode checkpoint` 会在 clean preflight 后创建或复用受 transaction record 管理的 `agent/<feature-id>` 本地分支。成功任务只有在 changed/untracked/sensitive 与 TaskSpec allowed/forbidden audit 通过后才 commit;失败任务的 patch、inventory、untracked content 与 boundary audit 写入 `.harness/task-pool/artifacts/features/<feature-id>/git/failures/`,随后 reset/clean 到最近 checkpoint 并再次验证 clean。默认模式不触碰 Git;`--keep-failed-diff` 只保留现场并停止,不允许继续下一个写任务。任何模式都不授权 stash、push、remote PR 或 merge。
95
-
96
- 业务 Task 失败不会跳过报告与最终 review,但命令返回 `needs-action` 和非零退出码;`noReadyReason` 用于区分 Closed、Deliverable、AwaitingQA、NeedsAction、dependency blocked 与 empty。
97
-
98
- ## Versioned Self-hosting Canary
99
-
100
- 源码仓库使用下面的 repo-maintainer 命令验证隔离 candidate package 的确定性接棒:
101
-
102
- ```bash
103
- npm run self-host:canary -- --deterministic --output <evidence.json>
104
- npm run self-host:canary -- --deterministic --tarball <candidate.tgz> --output <evidence.json>
105
- ```
106
-
107
- canary 会记录 tarball SHA-256、slot containment、两个 bin identity、由维护脚本独立计算并与候选实现交叉核对的 package fingerprint、每个 required check、skill mirror hashes、Feature dry-run、static/shell DAG run/report 以及 PATH trap。所有子进程都有硬超时;它删除 `HARNESS_ALLOW_ACTIVE_DAG_RUNS`,并要求 trap 中没有 `loop-agent`、`agent-worker` 或 `pi` 调用。`--live` 当前明确失败;deterministic canary 不包含 Pi/model/provider/cost 证明。
108
-
109
- ## Follow-up 决策与批准
110
-
111
- `agent-worker task draft-followup <task-id> --worker-run-id <id> --feature-dir <feature-dir> --repo <target-repo>` 从失败 run、TaskSpec 和 task graph 生成 `FU-<parent>-NN` 决策产物。ProductBug/TestBug/FlakyTest/DependencyFailure 生成可批准 TaskDraft;连续两次 EnvFailure 生成 `ENV-CHECK-*`,单次只建议 retry;SpecUnclear/ContractMismatch/RiskyChange/NeedsHuman/Unknown 生成人工行动卡。产物与 evidence/index 位于 `.harness/task-pool/artifacts/features/<feature-id>/follow-ups/`;生成过程受 Feature lock 和 dedupe key 保护,不写 Feature Packet、Ready state 或原失败事实。新 run 会 supersede 同 parent 的旧 unresolved 产物,但不会删除历史。
112
-
113
- 人工批准只接受 TaskDraft;行动卡不能批准或进入 Ready。批准会校验 draft/evidence hash、repo/path containment、Feature baseline 和冲突,在 staging validation 后写 TaskSpec、graph、Ready、approval/index 与 event。rename、state、approval、index、event 五个故障点均有注入回滚测试;重复批准幂等复用首次结果。`--dry-run` 零写入,`--json` 失败保持 stdout JSON、stderr 无噪声和非零退出码。
114
-
115
- ## Worker Retry
116
-
117
- 失败 Task Pool task 必须显式 retry,不能删除 `.harness/task-pool/runs.jsonl` 或复用失败的 `workerRunId`。canonical 命令必须携带 Feature:
118
-
119
- ```bash
120
- agent-worker task retry <task-id> --feature-id <feature-id> --repo <target-repo> --reason "provider configuration corrected"
121
- agent-worker batch run-ready --feature-dir <feature-dir> --repo <target-repo>
122
- ```
123
-
124
- `task retry` 只接受指定 Feature 下的 `Failed` state,保留旧 run 和 failure handoff,并写入 `retryOfWorkerRunId`。跨 Feature 同名 Task 时省略 `--feature-id` 必须 fail-closed。下一次 `run-ready` 生成新的 `workerRunId`;`Blocked`、`Done` 或无 state 的 task 必须先由 operator 处理根因,不能盲目重试。
125
-
126
- ## Task Pool Doctor 与 State 迁移
127
-
128
- ```bash
129
- agent-worker pool doctor --repo <target-repo> --json
130
- agent-worker pool migrate-state --repo <target-repo> # dry-run 默认
131
- agent-worker pool migrate-state --repo <target-repo> --apply --owner <owner> --reason <reason>
132
- ```
133
-
134
- `pool doctor` 只读扫描 v2 / legacy inventory 与映射证据,exit 0 也可带 findings。`pool migrate-state` 默认零写入;apply 需要 `--owner` 与 `--reason`,保留 bytes/hash 审计 artifact,失败全回滚,且不改写 `runs.jsonl` / `events.jsonl`。
135
-
136
- ## 验证 Preset
137
-
138
- Shell preset `loop-agent-standard-verify` 跑 typecheck 与聚焦 DAG 测试。完整交付仍应使用:
139
-
140
- ```bash
141
- bash scripts/ci.sh
142
- ```
@@ -1,96 +0,0 @@
1
- # Production Readiness v0.1
2
-
3
- 本文档冻结 2026-07-06 至 2026-07-12 hardening sprint 的单任务 production readiness 标准。
4
-
5
- `loop-agent` production readiness v0.1 指:低/中风险的仓库本地 DAG 任务能以稳定、可解释、可验证、可恢复的方式运行。这不是组织级平台 readiness 声明。
6
-
7
- ## 支持范围
8
-
9
- - 单仓库
10
- - 单任务或小范围有边界任务
11
- - 低/中风险的代码与文档变更
12
- - 显式 task source
13
- - 显式 `allowedPaths`、`forbiddenPaths`、`writeSet`
14
- - shell 验证命令
15
- - 通过 `dag report`、`dag doctor`、`promote-run`、`closeout` 交接
16
-
17
- ## 非目标
18
-
19
- - 自动 merge
20
- - 自动 release
21
- - 生产 secrets
22
- - 生产数据库访问
23
- - 高风险 migration
24
- - 在线 Worker Pool
25
- - 多仓库 feature 编排
26
- - 可写的 Dynamic Workflow sharded migration
27
- - DAG runtime 之外的第二套 runner
28
-
29
- ## 必需证据
30
-
31
- 每个 production-ready v0.1 任务必须保留或产出以下证据:
32
-
33
- | 证据 | 要求 |
34
- |---|---|
35
- | DAG spec path | 生成的 DAG JSON 路径记录在 CLI 输出、report 或 task artifacts 中 |
36
- | DAG validation result | `dag validate --strict-models --strict-governance` 有新鲜输出 |
37
- | Run id | `run-dag` 打印或记录 run id |
38
- | Shell verification output | 完成声明引用新鲜命令输出,而非 model 自报 |
39
- | Failure category | 失败 run 在可用时具备 raw、normalized、product-line 与 recommended follow-up 字段 |
40
- | Closeout 或 failure handoff | 成功 run 可 promote 并 closeout;失败 run 产出 failure handoff 证据而非成功 closeout |
41
-
42
- ## 必需命令
43
-
44
- 标准任务路径:
45
-
46
- ```bash
47
- loop-agent new-task <task-id> "Task title"
48
- loop-agent dag run-task <task-id> --profile auto --strict-models --output <temp-dir>/<task-id>-dag.json
49
- loop-agent dag validate --dag <temp-dir>/<task-id>-dag.json --strict-models --strict-governance
50
- loop-agent run-dag --dag <temp-dir>/<task-id>-dag.json --cwd .
51
- loop-agent dag report --run-id <run-id> --markdown
52
- loop-agent dag doctor --run-id <run-id>
53
- loop-agent promote-run <task-id> --run-id <run-id>
54
- loop-agent closeout task <task-id>
55
- ```
56
-
57
- `<temp-dir>` 使用平台原生临时目录。
58
-
59
- ## 必需标准
60
-
61
- | ID | 标准 | 验收 |
62
- |---|---|---|
63
- | PRD-01 | 新任务仅推荐 DAG 路径 | README、docs、website、skills 不把已移除的 Level 1 顺序命令当作 fallback |
64
- | PRD-02 | 写任务有显式边界 | strict DAG governance 在需要时拒绝缺失 `allowedPaths`、`forbiddenPaths` 或 `writeSet` |
65
- | PRD-03 | 每次 DAG run 有可读 report | `loop-agent dag report --latest --markdown` 显示 status、failure summary 与 next step |
66
- | PRD-04 | 失败 run 可诊断 | `loop-agent dag doctor` 报告 category 与 recommended follow-up |
67
- | PRD-05 | Shell 验证是完成权威 | closeout 不接受 model 自报替代新鲜命令输出 |
68
- | PRD-06 | Failure category 覆盖失败 fixture | validation、write guard、shell、timeout、auth、unknown fixture 均有映射 |
69
- | PRD-07 | 已完成 DAG facts 不可变 | completed facts guard 仍由测试覆盖 |
70
- | PRD-08 | CLI 主路径输出指引下一步 | `dag run-task`、`dag validate`、`run-dag`、`dag report` 打印 operator next commands |
71
- | PRD-09 | 文档面一致 | doc 更新后 `bash scripts/check-repo.sh` 与 `npm run docs:build` 通过 |
72
- | PRD-10 | sprint 完成前完整本地门禁通过 | 最终 hardening closeout 时 `bash scripts/ci.sh` 通过 |
73
-
74
- ## Failure Routing
75
-
76
- Failure category 是路由字段,不是主任务状态。保留 raw DAG facts,仅在 report、doctor 输出、closeout draft、task artifacts 或后续 Task Pool 记录中追加派生字段。
77
-
78
- 四层结构:
79
-
80
- ```text
81
- raw_failure_category
82
- dag_normalized_failure_category
83
- product_line_failure_category
84
- recommended_follow_up
85
- ```
86
-
87
- Product-line category 事实源:`ai_workspace/loop-agent/design/state-and-failure-taxonomy.md`。
88
-
89
- ## Sprint Gate
90
-
91
- 2026-07-06 sprint 仅在 active execution plan 记录以下内容时视为完成:
92
-
93
- - M1–M4 最终状态
94
- - 必需验证门禁的命令输出
95
- - 成功、write-guard、shell-failure 场景的 dogfood 证据
96
- - 剩余风险与后续工作
@@ -1,81 +0,0 @@
1
- # 进度日志
2
-
3
- 本目录存放应超越聊天上下文的简洁 handoff 日志。
4
-
5
- **索引职责**:本文件维护本目录**全量文件列表**。`ai_workspace/loop-agent/README.md` 只链到本 README。
6
-
7
- ## 近期要点(导读)
8
-
9
- - [`2026-07-18-observe-ops-surface-and-rich-timeline.md`](2026-07-18-observe-ops-surface-and-rich-timeline.md) — Observe 运营面/执行面重排、检查器半屏与富过程时间线收口(已归档 completed)
10
- - [`2026-07-17-observe-dashboard-incremental-polling-handoff.md`](2026-07-17-observe-dashboard-incremental-polling-handoff.md) — Observe 总览重复 loading/全量 DOM 重建根因、DAG 契约和跨机器恢复步骤
11
- - [`2026-07-17-0.13.0-merge-handoff.md`](2026-07-17-0.13.0-merge-handoff.md) — 0.13.0 合并、修复、验证与 Windows 剩余问题跨机器交接
12
- - [`2026-07-17-source-target-structure-alignment.md`](2026-07-17-source-target-structure-alignment.md) — 源仓库 canonical layout 对齐 fresh init,并记录首轮 write-set audit 修订
13
- - [`2026-07-17-fe-test-dag-archive.md`](2026-07-17-fe-test-dag-archive.md) — FE-test RAG DAG MVP + PR #25 收口归档(residual:prompt 深度 / browser smoke / website)
14
- - [`2026-07-17-frontend-implementation-m4-m6.md`](2026-07-17-frontend-implementation-m4-m6.md) — frontend-implementation 风险路由、项目能力、修复合同与 package surface 收口
15
- - [`2026-07-17-frontend-implementation-m3-repair.md`](2026-07-17-frontend-implementation-m3-repair.md) — frontend-implementation repair contract 与验证闭环
16
- - [`2026-07-17-frontend-implementation-m2-writer-trace.md`](2026-07-17-frontend-implementation-m2-writer-trace.md) — frontend writer trace 与 verification trace 证据
17
- - [`2026-07-17-frontend-implementation-contract-v1.md`](2026-07-17-frontend-implementation-contract-v1.md) — frontend implementation contract v1
18
- - [`2026-07-17-frontend-implementation-m0-baseline.md`](2026-07-17-frontend-implementation-m0-baseline.md) — frontend-implementation M0 能力基线
19
- - [`2026-07-16-frontend-mock-node.md`](2026-07-16-frontend-mock-node.md) — 前端 DAG Mock 策略、生产安全与真实接口证据边界(已验证)
20
- - [`2026-07-16-frontend-demand-auto-routing.md`](2026-07-16-frontend-demand-auto-routing.md) — 默认前端需求自动进入前端专用 DAG,后端/混合/未知保守回退(已验证)
21
- - [`2026-07-16-ready-planner.md`](2026-07-16-ready-planner.md) — Priority-aware Ready Planner:M1 core + RP-02 Feature/report/Observe 同源投影(已验证)
22
- - [`2026-07-15-task-pool-v2-feature-scoped-task-identity.md`](2026-07-15-task-pool-v2-feature-scoped-task-identity.md) — Task Pool v2 复合身份:M1 store + M2 consumers + M3 Observe + M4 文档收口(已归档 completed)
23
- - [`2026-07-15-init-update-surface-coverage.md`](2026-07-15-init-update-surface-coverage.md) — 修复初始化 surface 升级漏检(自动发现 copied 文件、blockedTargetPaths fail-safe、drift gate)实施交接(已验证)
24
- - [`2026-07-15-readonly-node-retry.md`](2026-07-15-readonly-node-retry.md) — 只读 Pi DAG 节点安全重试(默认策略、attempt 证据、validation 门禁)实施交接(已验证)
25
- - [`2026-07-15-repair-artifact-gate-runtime-contract-implementation.md`](2026-07-15-repair-artifact-gate-runtime-contract-implementation.md) — 显式 repairNodeId、runtime contract preflight 与 run-owned controller identity 实施交接(已验证)
26
- - [`2026-07-14-observe-ui-r5-boundary-final.md`](2026-07-14-observe-ui-r5-boundary-final.md) — R5 轻量入口与显式模块边界 DAG closeout
27
- - [`2026-07-14-observe-ui-r5-remediation.md`](2026-07-14-observe-ui-r5-remediation.md) — R5 浏览器状态所有权修复 DAG closeout
28
- - [`2026-07-14-observe-ui-r5.md`](2026-07-14-observe-ui-r5.md) — Observe 前端模块化与体验收敛(已验证)
29
- - [`2026-07-14-observe-ui-r4.md`](2026-07-14-observe-ui-r4.md) — Observe 事件时间线、有界 JSONL 与投影健康(已验证)
30
- - [`2026-07-14-observe-ui-r3.md`](2026-07-14-observe-ui-r3.md) — Observe Feature 决策详情与对象关联导航(已验证)
31
- - [`2026-07-14-observe-ui-r2.md`](2026-07-14-observe-ui-r2.md) — Observe Batch / Worker Run 证据化详情收口
32
- - [`2026-07-14-init-canonical-layout-handoff.md`](2026-07-14-init-canonical-layout-handoff.md) — 目标项目初始化布局迁移到 `ai_workspace/loop-agent/` 与 `.agents/skills/` 的跨会话交接
33
- - [`2026-07-14-self-update-notifier-implementation.md`](2026-07-14-self-update-notifier-implementation.md) — loop-agent CLI 自更新提醒实现、TDD 验证和 bounded DAG fallback 交接
34
- - [`2026-07-14-docs-living-docs-calibration.md`](2026-07-14-docs-living-docs-calibration.md) — 活文档校准(Dynamic Workflow / 六个月 / ADR)
35
- - [`2026-07-14-docs-governance-hygiene.md`](2026-07-14-docs-governance-hygiene.md) — 文档治理与双树/索引收敛
36
- - [`2026-07-14-website-docs-ia-and-converge.md`](2026-07-14-website-docs-ia-and-converge.md) — Website 文档 IA 交接
37
- - [`2026-07-13-sdd-embedded-skills.md`](2026-07-13-sdd-embedded-skills.md) — 目标项目 repo-local SDD skills 的可选 DAG 节点嵌入、验证与基线风险交接
38
- - [`2026-07-13-dev-controller-001-implementation.md`](2026-07-13-dev-controller-001-implementation.md) — controller identity 实施
39
- - [`2026-07-13-dev-snapshot-001-implementation.md`](2026-07-13-dev-snapshot-001-implementation.md) — skill snapshot 实施
40
- - [`2026-07-13-dev-canary-001-implementation.md`](2026-07-13-dev-canary-001-implementation.md) — candidate canary 实施
41
-
42
- ## 全量列表(新→旧)
43
-
44
- - [`2026-07-17-observe-dashboard-incremental-polling-handoff.md`](2026-07-17-observe-dashboard-incremental-polling-handoff.md)
45
- - [`2026-07-17-0.13.0-merge-handoff.md`](2026-07-17-0.13.0-merge-handoff.md)
46
- - [`2026-07-17-source-target-structure-alignment.md`](2026-07-17-source-target-structure-alignment.md)
47
- - [`2026-07-17-fe-test-dag-archive.md`](2026-07-17-fe-test-dag-archive.md)
48
- - [`2026-07-17-frontend-implementation-m4-m6.md`](2026-07-17-frontend-implementation-m4-m6.md)
49
- - [`2026-07-17-frontend-implementation-m3-repair.md`](2026-07-17-frontend-implementation-m3-repair.md)
50
- - [`2026-07-17-frontend-implementation-m2-writer-trace.md`](2026-07-17-frontend-implementation-m2-writer-trace.md)
51
- - [`2026-07-17-frontend-implementation-contract-v1.md`](2026-07-17-frontend-implementation-contract-v1.md)
52
- - [`2026-07-17-frontend-implementation-m0-baseline.md`](2026-07-17-frontend-implementation-m0-baseline.md)
53
- - [`2026-07-16-frontend-mock-node.md`](2026-07-16-frontend-mock-node.md)
54
- - [`2026-07-16-frontend-demand-auto-routing.md`](2026-07-16-frontend-demand-auto-routing.md)
55
- - [`2026-07-16-ready-planner.md`](2026-07-16-ready-planner.md)
56
- - [`2026-07-15-task-pool-v2-feature-scoped-task-identity.md`](2026-07-15-task-pool-v2-feature-scoped-task-identity.md)
57
- - [`2026-07-15-init-update-surface-coverage.md`](2026-07-15-init-update-surface-coverage.md)
58
- - [`2026-07-15-readonly-node-retry.md`](2026-07-15-readonly-node-retry.md)
59
- - [`2026-07-15-repair-artifact-gate-runtime-contract-implementation.md`](2026-07-15-repair-artifact-gate-runtime-contract-implementation.md)
60
- - [`2026-07-14-observe-ui-r5-boundary-final.md`](2026-07-14-observe-ui-r5-boundary-final.md)
61
- - [`2026-07-14-observe-ui-r5-remediation.md`](2026-07-14-observe-ui-r5-remediation.md)
62
- - [`2026-07-14-observe-ui-r5.md`](2026-07-14-observe-ui-r5.md)
63
- - [`2026-07-14-observe-ui-r4.md`](2026-07-14-observe-ui-r4.md)
64
- - [`2026-07-14-observe-ui-r3.md`](2026-07-14-observe-ui-r3.md)
65
- - [`2026-07-14-observe-ui-r2.md`](2026-07-14-observe-ui-r2.md)
66
- - [`2026-07-14-init-canonical-layout-handoff.md`](2026-07-14-init-canonical-layout-handoff.md)
67
- - [`2026-07-14-self-update-notifier-implementation.md`](2026-07-14-self-update-notifier-implementation.md)
68
- - [`2026-07-14-website-docs-ia-and-converge.md`](2026-07-14-website-docs-ia-and-converge.md)
69
- - [`2026-07-14-docs-living-docs-calibration.md`](2026-07-14-docs-living-docs-calibration.md)
70
- - [`2026-07-14-docs-governance-hygiene.md`](2026-07-14-docs-governance-hygiene.md)
71
- - [`2026-07-13-sdd-embedded-skills.md`](2026-07-13-sdd-embedded-skills.md)
72
- - [`2026-07-13-dev-snapshot-001-implementation.md`](2026-07-13-dev-snapshot-001-implementation.md)
73
- - [`2026-07-13-dev-controller-001-implementation.md`](2026-07-13-dev-controller-001-implementation.md)
74
- - [`2026-07-13-dev-canary-001-implementation.md`](2026-07-13-dev-canary-001-implementation.md)
75
- - [`2026-07-12-pi-only-agent-runtime.md`](2026-07-12-pi-only-agent-runtime.md)
76
- - [`2026-07-11-observe-dag-visualization.md`](2026-07-11-observe-dag-visualization.md)
77
- - [`2026-07-11-m2-feature-delivery-loop.md`](2026-07-11-m2-feature-delivery-loop.md)
78
- - [`2026-07-10-observe-ui-optimization.md`](2026-07-10-observe-ui-optimization.md)
79
- - [`2026-07-04-dynamic-workflow-dag-engine.md`](2026-07-04-dynamic-workflow-dag-engine.md)
80
- - [`2026-07-04-commander-cli-migration.md`](2026-07-04-commander-cli-migration.md)
81
- - [`2026-07-02-loop-agent-subject-restructure.md`](2026-07-02-loop-agent-subject-restructure.md)
@@ -1,163 +0,0 @@
1
- # 报告
2
-
3
- ## 合并报告命名
4
-
5
- 跨分支合并报告使用 `YYYY-MM-DD-origin-<source>-into-<target>-<source-short-sha>.md`。同一天重复合并时由 source short SHA 唯一标识;报告应记录双方功能、冲突文件与解析、自动合并高风险点、初始化/发布面影响、验证证据及剩余风险。新报告从 [`../templates/branch-merge-report.md`](../templates/branch-merge-report.md) 创建。
6
-
7
- - [`2026-07-18-origin-0.13.0-into-main-7c1a139.md`](2026-07-18-origin-0.13.0-into-main-7c1a139.md) — 将已发布并验证的 0.13.0 显式回合 main,合并后运行时代码树与发布分支一致
8
- - [`2026-07-18-origin-0.13.0-into-main-7c1a139-init-evolution-review.md`](2026-07-18-origin-0.13.0-into-main-7c1a139-init-evolution-review.md) — 0.13.0 回合 main 的 init/package/update surface 审查
9
- - [`2026-07-18-origin-main-into-0.13.0-43e86ad.md`](2026-07-18-origin-main-into-0.13.0-43e86ad.md) — 合入 frontend implementation M0–M6 与 browser-tools,同时保留 Observe、init/update 和既有受治理 DAG 能力
10
- - [`2026-07-18-origin-main-43e86ad-init-evolution-review.md`](2026-07-18-origin-main-43e86ad-init-evolution-review.md) — 本次合并的 package/full-init/update surface 审查
11
- - [`2026-07-17-origin-main-into-0.13.0-c139cf5.md`](2026-07-17-origin-main-into-0.13.0-c139cf5.md) — 合并 backend-test M0–M4、frontend-test/Playwright CLI 与 Pi SDK 更新,同时保留 0.13.0 初始化、Observe 和 Windows 稳定性修复
12
- - [`2026-07-17-origin-main-c139cf5-init-evolution-review.md`](2026-07-17-origin-main-c139cf5-init-evolution-review.md) — 本次高影响 merge 的 init/package/full-init/update 审查
13
-
14
- 本目录存放验证、审计、benchmark、init evolution 与 closeout 报告。
15
-
16
- **索引职责**:本文件维护本目录**全量文件列表**。`ai_workspace/loop-agent/README.md` 只链到本 README,不再精选单篇 report。
17
-
18
- 查找约定:
19
-
20
- - 按日期前缀浏览下方列表(新→旧)
21
- - init surface 审查:文件名含 `init-evolution-review`
22
- - 能力基线滚动分析:[`2026-07-02-repository-analysis.md`](2026-07-02-repository-analysis.md)(**已冻结**;2026-07-15 最终再采样关闭 0.11.0;非第二 CHANGELOG)
23
-
24
- ## 近期要点(导读,非权威全集)
25
-
26
- - [`2026-07-17-origin-main-merge.md`](2026-07-17-origin-main-merge.md) — origin/main 合并、canonical 路径收敛与 partial_failed 闭环说明
27
- - [`2026-07-17-review-findings-remediation-init-evolution-review.md`](2026-07-17-review-findings-remediation-init-evolution-review.md) — 主干 review findings 初始化影响审查
28
- - [`2026-07-16-frontend-mock-node-init-evolution-review.md`](2026-07-16-frontend-mock-node-init-evolution-review.md) — 前端 Mock 节点初始化影响审查
29
- - [`2026-07-16-frontend-mock-data-workflow.md`](2026-07-16-frontend-mock-data-workflow.md) — 前端 Mock 数据工作流报告
30
- - [`2026-07-16-frontend-mock-data-workflow-init-evolution-review.md`](2026-07-16-frontend-mock-data-workflow-init-evolution-review.md) — 前端 Mock 数据工作流初始化影响审查
31
- - [`2026-07-16-frontend-demand-auto-routing-init-evolution-review.md`](2026-07-16-frontend-demand-auto-routing-init-evolution-review.md) — 前端需求自动路由初始化影响审查
32
- - [`2026-07-16-dag-source-binding-recovery-init-evolution-review.md`](2026-07-16-dag-source-binding-recovery-init-evolution-review.md) — DAG 任务源绑定初始化影响审查
33
- - [`2026-07-16-backend-test-run-owned-report-init-evolution-review.md`](2026-07-16-backend-test-run-owned-report-init-evolution-review.md) — backend-test run-owned 报告初始化影响审查
34
- - [`2026-07-17-source-target-structure-alignment-partial-failed.md`](2026-07-17-source-target-structure-alignment-partial-failed.md) — 首轮结构迁移在 writer 前被 write-set audit 正确阻断的原因与 r2 修订
35
- - [`2026-07-17-weekly-progress-summary-last-tue-to-today.md`](2026-07-17-weekly-progress-summary-last-tue-to-today.md) — 管理层阶段进展总结(07-07~07-17:0.6→0.12、Observe/Feature/Task Pool/Planner)
36
- - [`2026-07-17-backend-test-analysis-contract-init-evolution-review.md`](2026-07-17-backend-test-analysis-contract-init-evolution-review.md) — backend-test run-owned 结构化分析契约与 JSON Schema 初始化投影审查
37
- - [`current-capability-summary.md`](current-capability-summary.md) — **活**能力摘要(新进展写这里 / CHANGELOG / 新 report)
38
- - [`2026-07-16-0.13.0-init-evolution-review.md`](2026-07-16-0.13.0-init-evolution-review.md) — 0.12.0→0.13.0 init 更新连续性修复与 knowledge surface 审查(init update required)
39
- - [`2026-07-16-task-pool-priority-aware-ready-planner.md`](2026-07-16-task-pool-priority-aware-ready-planner.md) — Priority-aware Ready Planner 关闭报告
40
- - [`2026-07-15-task-pool-v2-feature-scoped-task-identity.md`](2026-07-15-task-pool-v2-feature-scoped-task-identity.md) — Task Pool v2 复合身份 M0–M4 最终报告
41
- - [`2026-07-15-task-pool-v2-init-evolution-review.md`](2026-07-15-task-pool-v2-init-evolution-review.md) — Task Pool v2 文档变更的 init surface 非影响审查
42
- - [`2026-07-15-init-update-surface-coverage-init-evolution-review.md`](2026-07-15-init-update-surface-coverage-init-evolution-review.md) — 初始化 surface 升级漏检修复的 init surface 审查(init update required)
43
- - [`2026-07-15-readonly-node-retry-init-evolution-review.md`](2026-07-15-readonly-node-retry-init-evolution-review.md) — 只读 Pi DAG 节点安全重试的 init surface 审查(surface check only,无需 init evolution)
44
- - [`2026-07-15-backend-test-dag-init-evolution-review.md`](2026-07-15-backend-test-dag-init-evolution-review.md) — 后端测试专用 DAG 的 package/init surface 审查
45
- - [`2026-07-15-init-canonical-layout-init-evolution-review.md`](2026-07-15-init-canonical-layout-init-evolution-review.md) — init canonical layout、旧布局安全迁移和废弃 harness 字段清理审查
46
- - [`2026-07-14-self-update-notifier-implementation.md`](2026-07-14-self-update-notifier-implementation.md) — loop-agent CLI 自更新提醒实现、DAG 边界、TDD 与 init surface 非影响证据
47
- - [`2026-07-14-docs-living-docs-calibration.md`](2026-07-14-docs-living-docs-calibration.md) — 活文档校准(Dynamic Workflow / 六个月 / analysis 冻结 / ADR)
48
- - [`2026-07-14-docs-governance-hygiene.md`](2026-07-14-docs-governance-hygiene.md) — 文档治理扫描与索引/双树收敛
49
- - [`2026-07-14-website-docs-ia-and-converge.md`](2026-07-14-website-docs-ia-and-converge.md) — Website IA 与 docs-converge 收口
50
- - [`2026-07-13-sdd-embedded-skills.md`](2026-07-13-sdd-embedded-skills.md) — repo-local SDD skill 嵌入增强的 DAG、TDD、独立审查、验证和基线风险证据
51
- - [`2026-07-13-versioned-self-hosting-bootstrap.md`](2026-07-13-versioned-self-hosting-bootstrap.md) — 版本化自举证据
52
- - [`2026-07-12-m2-completion-audit.md`](2026-07-12-m2-completion-audit.md) — 第二月完成审计
53
- - [`2026-07-02-repository-analysis.md`](2026-07-02-repository-analysis.md) — **已冻结**仓库能力基线快照(2026-07-15 最终再采样至 0.11.0;勿再当 CHANGELOG)
54
-
55
- ## 全量列表(新→旧)
56
-
57
- - [`2026-07-18-origin-0.13.0-into-main-7c1a139.md`](2026-07-18-origin-0.13.0-into-main-7c1a139.md)
58
- - [`2026-07-18-origin-0.13.0-into-main-7c1a139-init-evolution-review.md`](2026-07-18-origin-0.13.0-into-main-7c1a139-init-evolution-review.md)
59
- - [`2026-07-18-origin-main-into-0.13.0-43e86ad.md`](2026-07-18-origin-main-into-0.13.0-43e86ad.md)
60
- - [`2026-07-18-origin-main-43e86ad-init-evolution-review.md`](2026-07-18-origin-main-43e86ad-init-evolution-review.md)
61
- - [`2026-07-17-pi-sdk-0-80-10-init-evolution-review.md`](2026-07-17-pi-sdk-0-80-10-init-evolution-review.md)
62
- - [`2026-07-17-origin-main-merge.md`](2026-07-17-origin-main-merge.md)
63
- - [`2026-07-17-init-evolution-review.md`](2026-07-17-init-evolution-review.md)
64
- - [`2026-07-17-frontend-implementation-m4-m6.md`](2026-07-17-frontend-implementation-m4-m6.md)
65
- - [`2026-07-17-frontend-implementation-m3-repair.md`](2026-07-17-frontend-implementation-m3-repair.md)
66
- - [`2026-07-17-frontend-implementation-m2-writer-trace.md`](2026-07-17-frontend-implementation-m2-writer-trace.md)
67
- - [`2026-07-17-frontend-implementation-m0-baseline.md`](2026-07-17-frontend-implementation-m0-baseline.md)
68
- - [`2026-07-17-frontend-implementation-contract-v1.md`](2026-07-17-frontend-implementation-contract-v1.md)
69
- - [`2026-07-17-backend-test-result-classification-m2.md`](2026-07-17-backend-test-result-classification-m2.md)
70
- - [`2026-07-17-backend-test-m4-converge-smoke.md`](2026-07-17-backend-test-m4-converge-smoke.md)
71
- - [`2026-07-17-backend-test-m0-template-runtime-drift.md`](2026-07-17-backend-test-m0-template-runtime-drift.md)
72
- - [`2026-07-17-backend-test-init-evolution-review.md`](2026-07-17-backend-test-init-evolution-review.md)
73
- - [`2026-07-17-backend-test-execution-contract-m1.md`](2026-07-17-backend-test-execution-contract-m1.md)
74
- - [`2026-07-17-backend-test-case-manifest-m3.md`](2026-07-17-backend-test-case-manifest-m3.md)
75
- - [`2026-07-17-source-target-structure-alignment-partial-failed.md`](2026-07-17-source-target-structure-alignment-partial-failed.md)
76
- - [`2026-07-16-0.13.0-init-evolution-review.md`](2026-07-16-0.13.0-init-evolution-review.md)
77
- - [`2026-07-17-weekly-progress-summary-last-tue-to-today.md`](2026-07-17-weekly-progress-summary-last-tue-to-today.md)
78
- - [`2026-07-17-backend-test-analysis-contract-artifact.md`](2026-07-17-backend-test-analysis-contract-artifact.md)
79
- - [`2026-07-17-backend-test-analysis-contract-init-evolution-review.md`](2026-07-17-backend-test-analysis-contract-init-evolution-review.md)
80
- - [`2026-07-17-review-findings-remediation-init-evolution-review.md`](2026-07-17-review-findings-remediation-init-evolution-review.md)
81
- - [`2026-07-16-dag-source-binding-recovery-init-evolution-review.md`](2026-07-16-dag-source-binding-recovery-init-evolution-review.md)
82
- - [`2026-07-16-repair-artifact-pass-normalization.md`](2026-07-16-repair-artifact-pass-normalization.md)
83
- - [`2026-07-16-observe-inspector-node-consistency.md`](2026-07-16-observe-inspector-node-consistency.md)
84
- - [`2026-07-16-init-natural-language-routing-init-evolution-review.md`](2026-07-16-init-natural-language-routing-init-evolution-review.md)
85
- - [`2026-07-16-decision-envelope-prompt-contract.md`](2026-07-16-decision-envelope-prompt-contract.md)
86
- - [`2026-07-16-decision-envelope-fence-tolerance.md`](2026-07-16-decision-envelope-fence-tolerance.md)
87
- - [`2026-07-16-frontend-mock-node-init-evolution-review.md`](2026-07-16-frontend-mock-node-init-evolution-review.md)
88
- - [`2026-07-16-frontend-demand-auto-routing-init-evolution-review.md`](2026-07-16-frontend-demand-auto-routing-init-evolution-review.md)
89
- - [`2026-07-16-frontend-mock-data-workflow.md`](2026-07-16-frontend-mock-data-workflow.md)
90
- - [`2026-07-16-frontend-mock-data-workflow-init-evolution-review.md`](2026-07-16-frontend-mock-data-workflow-init-evolution-review.md)
91
- - [`2026-07-16-task-pool-priority-aware-ready-planner.md`](2026-07-16-task-pool-priority-aware-ready-planner.md)
92
- - [`2026-07-15-task-pool-v2-feature-scoped-task-identity.md`](2026-07-15-task-pool-v2-feature-scoped-task-identity.md)
93
- - [`2026-07-15-task-pool-v2-init-evolution-review.md`](2026-07-15-task-pool-v2-init-evolution-review.md)
94
- - [`2026-07-15-init-update-surface-coverage-init-evolution-review.md`](2026-07-15-init-update-surface-coverage-init-evolution-review.md)
95
- - [`2026-07-15-readonly-node-retry-init-evolution-review.md`](2026-07-15-readonly-node-retry-init-evolution-review.md)
96
- - [`2026-07-15-backend-test-dag-init-evolution-review.md`](2026-07-15-backend-test-dag-init-evolution-review.md)
97
- - [`2026-07-15-init-canonical-layout-init-evolution-review.md`](2026-07-15-init-canonical-layout-init-evolution-review.md)
98
- - [`dogfood-preflight-static-20260709090443.md`](dogfood-preflight-static-20260709090443.md)
99
- - [`current-capability-summary.md`](current-capability-summary.md)
100
- - [`2026-07-14-website-docs-ia-init-evolution-review.md`](2026-07-14-website-docs-ia-init-evolution-review.md)
101
- - [`2026-07-14-observe-ui-r5.md`](2026-07-14-observe-ui-r5.md)
102
- - [`2026-07-14-observe-ui-r4.md`](2026-07-14-observe-ui-r4.md)
103
- - [`2026-07-14-observe-ui-r3.md`](2026-07-14-observe-ui-r3.md)
104
- - [`2026-07-14-observe-ui-r2.md`](2026-07-14-observe-ui-r2.md)
105
- - [`2026-07-14-observe-ui-r1.md`](2026-07-14-observe-ui-r1.md)
106
- - [`2026-07-14-architecture-docs-init-evolution-review.md`](2026-07-14-architecture-docs-init-evolution-review.md)
107
- - [`2026-07-14-website-docs-ia-and-converge.md`](2026-07-14-website-docs-ia-and-converge.md)
108
- - [`2026-07-14-self-update-notifier-implementation.md`](2026-07-14-self-update-notifier-implementation.md)
109
- - [`2026-07-14-docs-living-docs-calibration.md`](2026-07-14-docs-living-docs-calibration.md)
110
- - [`2026-07-14-docs-governance-hygiene.md`](2026-07-14-docs-governance-hygiene.md)
111
- - [`2026-07-13-versioned-self-hosting-init-evolution-review.md`](2026-07-13-versioned-self-hosting-init-evolution-review.md)
112
- - [`2026-07-13-versioned-self-hosting-bootstrap.md`](2026-07-13-versioned-self-hosting-bootstrap.md)
113
- - [`2026-07-13-test-environment-stability.md`](2026-07-13-test-environment-stability.md)
114
- - [`2026-07-13-sdd-embedded-skills-init-evolution-review.md`](2026-07-13-sdd-embedded-skills-init-evolution-review.md)
115
- - [`2026-07-13-sdd-embedded-skills.md`](2026-07-13-sdd-embedded-skills.md)
116
- - [`2026-07-13-frontend-verification-command-safety-init-evolution-review.md`](2026-07-13-frontend-verification-command-safety-init-evolution-review.md)
117
- - [`2026-07-13-frontend-skill-contracts-init-evolution-review.md`](2026-07-13-frontend-skill-contracts-init-evolution-review.md)
118
- - [`2026-07-13-exec-plan-lifecycle-init-evolution-review.md`](2026-07-13-exec-plan-lifecycle-init-evolution-review.md)
119
- - [`2026-07-13-exec-plan-lifecycle.md`](2026-07-13-exec-plan-lifecycle.md)
120
- - [`2026-07-12-production-readiness-hardening.md`](2026-07-12-production-readiness-hardening.md)
121
- - [`2026-07-12-pi-only-agent-runtime-init-evolution-review.md`](2026-07-12-pi-only-agent-runtime-init-evolution-review.md)
122
- - [`2026-07-12-observe-warm-console-redesign.md`](2026-07-12-observe-warm-console-redesign.md)
123
- - [`2026-07-12-observe-dag-node-model.md`](2026-07-12-observe-dag-node-model.md)
124
- - [`2026-07-12-observe-dag-node-model-init-evolution-review.md`](2026-07-12-observe-dag-node-model-init-evolution-review.md)
125
- - [`2026-07-12-m2-completion-audit.md`](2026-07-12-m2-completion-audit.md)
126
- - [`2026-07-12-m2-08-init-evolution-review.md`](2026-07-12-m2-08-init-evolution-review.md)
127
- - [`2026-07-12-m2-08-dogfood-release-evidence.md`](2026-07-12-m2-08-dogfood-release-evidence.md)
128
- - [`2026-07-12-m2-07-init-evolution-review.md`](2026-07-12-m2-07-init-evolution-review.md)
129
- - [`2026-07-12-m2-06-init-evolution-review.md`](2026-07-12-m2-06-init-evolution-review.md)
130
- - [`2026-07-12-m2-05-init-evolution-review.md`](2026-07-12-m2-05-init-evolution-review.md)
131
- - [`2026-07-12-m2-04-init-evolution-review.md`](2026-07-12-m2-04-init-evolution-review.md)
132
- - [`2026-07-12-m2-03-init-evolution-review.md`](2026-07-12-m2-03-init-evolution-review.md)
133
- - [`2026-07-12-m2-02-init-evolution-review.md`](2026-07-12-m2-02-init-evolution-review.md)
134
- - [`2026-07-12-init-evolution-strict-report-range.md`](2026-07-12-init-evolution-strict-report-range.md)
135
- - [`2026-07-12-dag-state-reconciliation.md`](2026-07-12-dag-state-reconciliation.md)
136
- - [`2026-07-12-dag-state-reconciliation-init-evolution-review.md`](2026-07-12-dag-state-reconciliation-init-evolution-review.md)
137
- - [`2026-07-12-dag-output-language.md`](2026-07-12-dag-output-language.md)
138
- - [`2026-07-12-dag-output-language-init-evolution-review.md`](2026-07-12-dag-output-language-init-evolution-review.md)
139
- - [`2026-07-12-dag-liveness-recovery.md`](2026-07-12-dag-liveness-recovery.md)
140
- - [`2026-07-12-dag-liveness-init-evolution-review.md`](2026-07-12-dag-liveness-init-evolution-review.md)
141
- - [`2026-07-12-0.10.0-init-evolution-review.md`](2026-07-12-0.10.0-init-evolution-review.md)
142
- - [`2026-07-11-task-pool-hard-cutover-init-evolution-review.md`](2026-07-11-task-pool-hard-cutover-init-evolution-review.md)
143
- - [`2026-07-11-round3-worker-expansion.md`](2026-07-11-round3-worker-expansion.md)
144
- - [`2026-07-11-round2-interactive-ui-experiment.md`](2026-07-11-round2-interactive-ui-experiment.md)
145
- - [`2026-07-11-round2-init-evolution-review.md`](2026-07-11-round2-init-evolution-review.md)
146
- - [`2026-07-11-observe-terminal-dag-kpi.md`](2026-07-11-observe-terminal-dag-kpi.md)
147
- - [`2026-07-11-observe-dag-visualization.md`](2026-07-11-observe-dag-visualization.md)
148
- - [`2026-07-11-m2-01-init-evolution-review.md`](2026-07-11-m2-01-init-evolution-review.md)
149
- - [`2026-07-11-init-evolution-review.md`](2026-07-11-init-evolution-review.md)
150
- - [`2026-07-11-commit-1d0f3b7-review.md`](2026-07-11-commit-1d0f3b7-review.md)
151
- - [`2026-07-11-command-performance-followups.md`](2026-07-11-command-performance-followups.md)
152
- - [`2026-07-11-command-performance-audit.md`](2026-07-11-command-performance-audit.md)
153
- - [`2026-07-11-0.7.5-init-evolution-review.md`](2026-07-11-0.7.5-init-evolution-review.md)
154
- - [`2026-07-10-session-events-dogfood.md`](2026-07-10-session-events-dogfood.md)
155
- - [`2026-07-10-nightly-drill-round-1.md`](2026-07-10-nightly-drill-round-1.md)
156
- - [`2026-07-10-nightly-drill-round-1-evidence.md`](2026-07-10-nightly-drill-round-1-evidence.md)
157
- - [`2026-07-10-next-stage-dogfood-runs.md`](2026-07-10-next-stage-dogfood-runs.md)
158
- - [`2026-07-10-init-evolution-review.md`](2026-07-10-init-evolution-review.md)
159
- - [`2026-07-09-be-001-pi-real-run.md`](2026-07-09-be-001-pi-real-run.md)
160
- - [`2026-07-05-website-docs-governance-multi-executor-dogfood.md`](2026-07-05-website-docs-governance-multi-executor-dogfood.md)
161
- - [`2026-07-04-website-docs-dynamic-workflow-dogfood.md`](2026-07-04-website-docs-dynamic-workflow-dogfood.md)
162
- - [`2026-07-02-repository-analysis.md`](2026-07-02-repository-analysis.md)
163
- - [`2026-06-06-pi-sdk-runtime-reuse-benchmark.md`](2026-06-06-pi-sdk-runtime-reuse-benchmark.md)
@@ -1,70 +0,0 @@
1
- # 验证矩阵
2
-
3
- 用能证明声明的最窄命令。
4
-
5
- | 变更类型 | 最低验证 | 更强验证 |
6
- |---|---|---|
7
- | 仅文档或治理 | `bash scripts/check-repo.sh` | `bash scripts/ci.sh` |
8
- | TypeScript runtime | `npm run typecheck` | `npm test` |
9
- | CLI 行为 | `npm run typecheck` + 定向 Vitest | `npm test` + CLI smoke |
10
- | Feature reducer/review | `npx vitest run test/worker/feature test/worker/agent-worker-cli.test.ts` | `npm test` + `node bin/agent-worker.js feature review --help` |
11
- | Feature run orchestration | `npx vitest run test/worker/feature test/worker/runner/run-ready.test.ts test/worker/agent-worker-cli.test.ts` | `npm test` + `node bin/agent-worker.js feature run --help` |
12
- | Follow-up 分类与事务 | `npx vitest run test/worker/follow-up test/worker/feature test/worker/pool test/worker/report test/worker/agent-worker-cli.test.ts` | `npm test` + `node bin/agent-worker.js task draft-followup --help` + `node bin/agent-worker.js feature approve-followup --help` |
13
- | Feature Git transaction | `npx vitest run test/worker/delivery test/worker/feature test/worker/runner test/worker/agent-worker-cli.test.ts` | `npm test` + `node bin/agent-worker.js feature run --help` + temporary Git repo smoke |
14
- | Final verification / Delivery / Closeout | `npx vitest run test/worker/delivery test/worker/feature test/worker/agent-worker-cli.test.ts` | `npm test` + verify-final/Delivery/Closeout CLI help + temporary Git repo evidence/hash/tamper/apply rollback smoke。Closeout 覆盖在 `test/worker/delivery/**` 与 CLI,不存在独立 `test/worker/closeout/` 目录 |
15
- | Task Pool feature-scoped identity / migration | `npx vitest run test/worker/pool test/worker/runner/run-ready.test.ts test/worker/feature test/worker/follow-up test/worker/delivery test/worker/report test/worker/observability test/worker/observe test/worker/agent-worker-cli.test.ts` + `npm run typecheck` | 平台临时目录 dual-Feature smoke(同名 `QA-EXEC-001` 不串线)+ `pool doctor` / `pool migrate-state` dry-run + Observe canonical/legacy 409 |
16
- | Controller identity / pinning | `npx vitest run test/worker/package-metadata.test.ts test/worker/repo-preflight.test.ts test/worker/loop-agent-client.test.ts test/worker/run-task/run-task.test.ts test/worker/runner/run-ready.test.ts test/worker/feature/run.test.ts test/worker/delivery/final-verification.test.ts test/worker/delivery/package.test.ts test/worker/agent-worker-cli.test.ts` + `npm run typecheck` | `npm test` + build + 两个 CLI `--version` + 写入型命令 help |
17
- | DAG resolved skill snapshot / resume | `npx vitest run test/dag-skill-snapshot.test.ts test/dag-approve-resume.test.ts` + `npm run typecheck` | `npm test` + snapshot tamper/pause-resume/dynamic profile smoke |
18
- | Versioned self-hosting candidate canary | `npx vitest run test/self-host-canary.test.ts test/package-surface.test.ts` + `npm run build` | `npm run self-host:canary -- --deterministic --output <evidence.json>` + `npm pack --dry-run` |
19
- | Public skill / init mirror surface | `bash scripts/check-skill-entry.sh` + `npx vitest run test/init-command.test.ts test/package-surface.test.ts` | 临时目标项目 full init + skill/mirror hash 对照 + `bash scripts/check-repo.sh` |
20
- | DAG 工作流 | 定向 DAG 测试 | `npm test` |
21
- | Eval Replay MVP | `npx vitest run test/eval-replay.test.ts test/eval-cli.test.ts test/cli-contract.test.ts` + `npm run typecheck` | `bash scripts/check-command-registry-drift.sh` + `bash scripts/check-architecture-boundaries.sh` + `node bin/loop-agent.js eval --help` + `bash scripts/check-repo.sh` |
22
- | Eval Candidate Registry MVP | `npx vitest run test/eval-candidate-registry.test.ts test/eval-cli.test.ts test/cli-contract.test.ts` + `npm run typecheck` | `npm run build` + `bash scripts/check-command-registry-drift.sh` + `bash scripts/check-architecture-boundaries.sh` + `node bin/loop-agent.js eval --help` + `HARNESS_ALLOW_ACTIVE_DAG_RUNS=1 bash scripts/check-repo.sh` |
23
- | Eval run association W2.3 | `npx vitest run test/eval-run-association.test.ts test/eval-replay.test.ts test/dag-report.test.ts test/run-dag.test.ts` + `npm run typecheck` | `npm run build` + `bash scripts/check-architecture-boundaries.sh` + `bash scripts/check-command-registry-drift.sh` + `HARNESS_ALLOW_ACTIVE_DAG_RUNS=1 bash scripts/check-repo.sh` + `npm run docs:build` + `git diff --check` |
24
- | 前端 DAG 模板 / 前端节点生成 | `npm run typecheck` + 定向 DAG 测试 | `npm test` + `bash scripts/check-repo.sh` |
25
- | Production readiness hardening | `bash scripts/check-repo.sh` + 定向 DAG/CLI 测试 | `bash scripts/ci.sh` + docs build + package smoke |
26
- | 脚本或 CI | 运行变更的脚本 | `bash scripts/ci.sh` |
27
- | Package / publish 入口 | `npm run build` + `node bin/loop-agent.js --help` | `npm pack --dry-run` |
28
- | 完整交付 | `bash scripts/ci.sh` | CLI smoke + 相关手工检查 |
29
-
30
- 常用命令:
31
-
32
- ```bash
33
- npm run typecheck
34
- npm test
35
- npm run build
36
- bash scripts/check-repo.sh
37
- bash scripts/ci.sh
38
- node bin/loop-agent.js --help
39
- node bin/loop-agent.js --version
40
- node bin/agent-worker.js --version
41
- npm run dev -- --help
42
- npm pack --dry-run
43
- npm run self-host:canary -- --deterministic --output <evidence.json>
44
- ```
45
-
46
- deterministic self-host canary 必须只执行 static/shell DAG,并在 evidence 中证明 `piExecutorObserved=false`、`modelExecutorObserved=false`、`featureExecutedTasks=[]`、PATH controller trap 未触发且候选入口位于隔离 slot。它不证明 candidate Pi/model/provider 路径,也不替代 `test/dag-skill-snapshot.test.ts` 对候选 skill resolution、snapshot hash 和 resume 的验证。
47
-
48
- 测试资源策略:`npm test` 会先运行可并行的纯测试,再串行运行会启动 CLI、`tsx` 或 Git 子进程的集成测试。纯测试根据 `os.availableParallelism()` 自适应使用 1 至 4 个 Worker;可用 `LOOP_AGENT_FAST_TEST_WORKERS` 在 CI 或受限机器上显式覆盖。传入测试文件时,例如 `npm test -- test/worker/observe/server.test.ts`,仍按常规 Vitest 方式只运行该范围,不会启动两阶段完整测试。
49
-
50
- 产品线 Feature/Task/QA 文档或 TaskSpec 约束变更还应运行:
51
-
52
- ```bash
53
- bash scripts/check-product-line-docs.sh
54
- ```
55
-
56
- 该检查对仓库中的 feature packet 执行 `agent-worker task validate-feature`,覆盖 AC 唯一性、依赖/环、TaskSpec 路径与验证命令、QA verdict 与 success closeout 顺序。
57
-
58
- Windows 上通过 Git Bash 或已配置的兼容 Bash 运行 `scripts/*.sh`。跨平台 CLI 代码对真实文件操作用原生路径;`/` 仅用于稳定 repo 引用、JSON/Markdown 证据引用和 glob 约定。
59
-
60
- 没有相关门禁的新鲜命令输出,不得声明完成。
61
-
62
- Production Readiness v0.1 工作以 `ai_workspace/loop-agent/production-readiness.md` 与 `ai_workspace/loop-agent/templates/production-readiness-checklist.md` 为验收契约。最终 hardening closeout 需要:
63
-
64
- ```bash
65
- bash scripts/ci.sh
66
- npm run docs:build
67
- npm run build
68
- node bin/loop-agent.js --help
69
- npm pack --dry-run
70
- ```
@@ -1,29 +0,0 @@
1
- #!/usr/bin/env bash
2
- set -euo pipefail
3
-
4
- ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
5
- cd "${ROOT_DIR}"
6
-
7
- feature_dirs=()
8
- for root in features product/features dogfood/features; do
9
- [[ -d "${root}" ]] || continue
10
- while IFS= read -r feature_dir; do
11
- if [[ ! -f "${feature_dir}/acceptance.yaml" ]]; then
12
- echo "product-line docs check failed: incomplete feature packet missing acceptance.yaml: ${feature_dir}" >&2
13
- exit 1
14
- fi
15
- feature_dirs+=("${feature_dir}")
16
- done < <(find "${root}" -mindepth 1 -maxdepth 1 -type d | sort)
17
- done
18
-
19
- if [[ ${#feature_dirs[@]} -eq 0 ]]; then
20
- echo "product-line docs check skipped: no feature packets found"
21
- exit 0
22
- fi
23
-
24
- for feature_dir in "${feature_dirs[@]}"; do
25
- echo "==> validate product-line feature: ${feature_dir}"
26
- node --import tsx/esm src/worker/cli.ts task validate-feature "${feature_dir}"
27
- done
28
-
29
- echo "product-line docs checks passed: ${#feature_dirs[@]} feature packet(s)"