@mstar-harness/opencode 1.0.0 → 1.0.2

package/CHANGELOG.md

@@ -0,0 +1,254 @@
+# Changelog
+
+All notable changes to the `@mstar-harness/opencode` package are documented in this file.
+
+The monorepo root [CHANGELOG.md](../../CHANGELOG.md) summarizes cross-surface releases.
+
+## 1.0.2
+
+- Bundled skills/commands: iteration artifact boundaries (`{ITERATION_DIR}/<id>/` workspace, specs-first start, compound workspace promotion at close); new iteration references; corpus hygiene aligned.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **1.0.2**.
+
+## 1.0.1
+
+- Bundled skills/commands: `iteration-drive` + `mstar-iteration` Phase 2 per-task SDD mandate; optional sticky implementer session (`resume` + ledger); thin `pm` skill shim; routing-eval v14.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **1.0.1**.
+
+## 1.0.0
+
+- Bundle `mstar-sdd` skill + scripts; SDD mandatory plan QC tri-review (`Execution mode: sdd`); inline/hotfix single-seat; plan template and SDD metadata fields.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **1.0.0**.
+
+## 0.7.9
+
+- Bundled commands/skills: Assignment plain role ids across PM templates, dispatch gates, branch-worktree, iteration commands, opencode host hygiene, and role NEVER rules — avoids OpenCode `@` auto-dispatch in harness prose.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.9**.
+
+## 0.7.8
+
+- Bundled commands/skills: `mstar-iteration` Phase 4–5 (PR delivery + merge-ready loop); `iteration-drive` sequences through Phase 5 with optional greploop/babysit helpers; core index and anti-patterns updated.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.8**.
+
+## 0.7.7
+
+- Bundled commands/skills: standalone `mstar-*` invariant; bundle `grill-me` for `/iteration-start` only; remove Context7/OpenViking/Open Design from runtime load paths; distill `open-harness-principles.md` into core and plan-conventions references; keep Role → typical topic skills matrix in `mstar-roles`.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.7**.
+
+## 0.7.6
+
+- Bundled commands/skills: commands–skills layering for iteration (`iteration-start` / `iteration-drive` orchestration vs `mstar-iteration` / `mstar-dispatch-gates` SSOT); hardened dispatch gates, Phase 2→3→PR transitions, canonical host dispatch (no Mode A/B/C fallbacks); Phase 1 Review & Edit chain is **sequential** product-manager → architect → writing-specialist.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.6**.
+
+## 0.7.5
+
+- Bundled commands/skills: explicit iteration branch policy — `iteration_base_branch`, `spec_integration_branch`, and `target_branch` required in compass/status metadata; no silent default to `main` for integration branch or PR target. Hardened `iteration-start` / `iteration-drive` gates and QC merge-base rules.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.5**.
+
+## 0.7.4
+
+- Bundled skills/docs: remove Superpowers runtime dependency from Morning Star guidance and host docs; assignments now rely on native mstar dispatch, worktree, plan, review, and evidence contracts.
+- Bundled skills: add `mstar-skill-authoring`, delete `mstar-execution-practices`, and fold review feedback handling into `mstar-coding-behavior`.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.4**.
+
+## 0.7.3
+
+- Bundled commands/skills: `mstar-iteration` Phase 3 close is now an explicit gate after all plans are `Done`; close requires compass normalization when needed, compound Phase 6 knowledge indexing, roadmap update, completed frontmatter, and close checklists before PR.
+- Bundled docs: README command table now includes `/mstar-bootstrap`; workflow and skill overview now match the iteration lifecycle.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.3**.
+
+## 0.7.2
+
+- Docs only at package level: root **0.7.2** documents CLI Cursor install layout (real git checkout at plugin path, not symlink). Re-bundle on publish if CLI dist changed.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.2**.
+
+## 0.7.1
+
+- Bundled commands/skills: `/iteration-start` §5 Review & Edit chain is a hard gate before integration branch commit — Task dispatch for product-manager, architect, writing-specialist; evidence is edited docs + locked compass (no iteration `reports/` artifacts). Synced in `mstar-iteration` §1.6, `pm`, `mstar-dispatch-gates`, `mstar-harness-core`.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.7.1**.
+
+## 0.6.22
+
+- Bundled skills: `mstar-dispatch-gates`, `mstar-roles` (qc-specialist-shared, dispatch-and-assignment) — replace prohibition-based anti-recursion with identity-deprivation framework. Assignment template gains IDENTITY + CAPABILITY BOUNDARY blocks before prohibitions, shifting from "I must not dispatch" (negation) to "I am a leaf executor; Task is not my tool" (capability deprivation).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.22**.
+
+## 0.6.21
+
+- Bundled skills: `mstar-design-md` — add YAML frontmatter as SSOT for token values in templates and spec. Bump template format version to 0.1.0.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.21**.
+
+## 0.6.20
+
+- Bundled commands: `/iteration-start` §5 changed to "Review & Edit Chain" — each reviewer now directly edits documents rather than only flagging issues. PM only does the final lock.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.20**.
+
+## 0.6.19
+
+- Bundled skills: `mstar-coding-behavior` strengthened with distilled Ponytail principles — YAGNI gate, The Ladder (7-level decision hierarchy), "deletion over addition / boring over clever", `simplify:` marker discipline, "bug fix = root cause", and "minimal check for non-trivial logic".
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.19**.
+
+## 0.6.18
+
+- Bundled commands: `/iteration-start` now includes explicit `## 0. Boot` section aligned with `/iteration-drive`, loading core harness entry and PM role identity before research.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.18**.
+
+## 0.6.17
+
+- Bundled commands: `/iteration-drive` PR target now resolved from iteration metadata (`target_branch` in `status.json`), defaulting to `main` when not set.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.17**.
+
+## 0.6.16
+
+- Bundled commands: Add `/iteration-drive` command that invokes the PM Autonomous Execute driver to push all non-`Done` plans to completion (implement → QC → QA → Done loop, then optional PR to `main`).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.16**.
+
+## 0.6.15
+
+- Bundled commands: New `harness-commands/` directory bundle and `loadBundledCommands()` registration in the config hook. Adds `/iteration-start` command that guides PM through iteration bootstrap (research, explore, lock, compass/plans, review chain, integration branch).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.15**.
+
+## 0.6.11
+
+- Bundled agents: Cursor-first frontmatter on all role shells so hosts that share `agents/*.md` parse `name`/`description`/`model` before OpenCode fields.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.11**.
+
+## 0.6.10
+
+- Bundled skills: Profile B `archived/plans-done.json` schema is `{ "plans": [<plan-id>, ...] }` only; add `plans-done.empty.json` template and bootstrap/PM init notes.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.10**.
+
+## 0.6.9
+
+- Bundled skills: universal `pm` orchestration entry (Cursor/Codex `/pm`, OpenCode PM role switch) and Autonomous Execute driver (`status.json` backlog, `spec_integration_branch`, per-plan feature branches, host todos).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.9**.
+
+## 0.6.8
+
+- Bundled skills: targeted QC re-review after fixes (owner seat only, in-place reports), short QC report basenames under `reports/<plan-id>/` (`qc1.md` …).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.8**.
+
+## 0.6.7
+
+- Bundled skills: Codex Plan / Goal Mode bridge for keeping `/plan`, `update_plan`, `/goal`, and goal progress aligned with `.mstar/` SSOT.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.7**.
+
+## 0.6.6
+
+- Bundled skills/agents: Codex custom-agent source files, `.mstar/` harness defaults, and updated host install docs.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.6**.
+
+## 0.6.5
+
+- Bundled skills: Durable Roadmap Gate for staged/partial/temporary work, plus routing-eval v8.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.5**.
+
+## 0.6.4
+
+- Bundled skills: Cursor Plan Build resume contract and routing-eval v7.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.4**.
+
+## 0.6.3
+
+- Bundled skills: slim `/pm` with dispatch-first + `/pm`-only rules; PM shell pointer.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.3**.
+
+## 0.6.2
+
+- Bundled skills: `/pm` Autonomous Execute push (iteration driver, multi-plan); PM shell cross-reference.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.2**.
+
+## 0.6.1
+
+- Bundled skills: `tech-debt-rollup.sh`, PM dual fullstack spread defaults, routing-eval v6.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.1**.
+
+## 0.6.0
+
+- Unified **`mstar-host`** in bundled `harness-skills/`; removed separate package `skills/` host path.
+
+See root [CHANGELOG.md](../../CHANGELOG.md) **0.6.0**.
+
+## 0.5.1
+
+### Bundled harness skills (`harness-skills/` at publish)
+
+- Cursor Plan mode dual-write bridge (`mstar-host`, `cursor-plan-mode-bridge`, `pm`, `mstar-harness-core`); `rules/mstar-cursor-plan-mode.mdc`.
+
+See root [CHANGELOG.md](../../CHANGELOG.md).
+
+## 0.5.0
+
+### Bundled harness skills (`harness-skills/` at publish)
+
+- Align bundled assets with the 0.5.0 harness release, including Codex plugin validation fixes in shared skill metadata.
+
+See root [CHANGELOG.md](../../CHANGELOG.md).
+
+## 0.4.1
+
+### Bundled harness skills (`harness-skills/` at publish)
+
+- **`mstar-plan-artifacts/templates/`**: `status.empty.json` and `notes.empty.json` moved from `mstar-plan-conventions` (paths in skill text updated).
+
+See root [CHANGELOG.md](../../CHANGELOG.md).
+
+## 0.4.0
+
+### Bundled harness skills (`harness-skills/` at publish)
+
+- **Topic skill split**: `mstar-phase-gates`, `mstar-dispatch-gates`, `mstar-branch-worktree`, `mstar-plan-artifacts` (status/residual included); slimmer `mstar-harness-core` and `mstar-plan-conventions`.
+- **`mstar-roles`**: Per-role required skill lists; host adapters updated for on-demand topic loading.
+- **`mstar-plan-conventions`**: `{ITERATION_DIR}`, `{KNOWLEDGE_DIR}`, content boundaries; optional `iteration_compass` / `iteration_refs`.
+- **`mstar-phase-gates`**: Prepare **`clarify` core discipline** (shared understanding, explore before asking, recommended answers).
+
+See root [CHANGELOG.md](../../CHANGELOG.md) for full release notes.
+
+## 0.3.2
+
+### Bundled harness skills (`harness-skills/` at publish)
+
+- **`mstar-plan-conventions`**: Formalize `{ITERATION_DIR}` (`{HARNESS_DIR}/iterations/`) and `{KNOWLEDGE_DIR}` (`{HARNESS_DIR}/knowledge/`); add `docs/` vs harness subtree content boundaries; optional `iteration_compass` / `iteration_refs` in `status.json` metadata.
+- **`mstar-harness-core`**: Embed Prepare **`clarify` core discipline** — walk the design decision tree to shared understanding, explore the codebase before asking the user, provide a recommended answer per question.
+
+## 0.3.1
+
+- Align package version with monorepo **0.3.1** (see root `CHANGELOG.md` for harness and host-adapter notes bundled via `harness-skills/` / `harness-agents/`).
+
+## 0.3.0
+
+- Align package version with monorepo **0.3.0** (see root `CHANGELOG.md` for harness and host-adapter notes bundled via `harness-skills/` / `harness-agents/`).
+
+## 0.2.0
+
+- Publish OpenCode plugin with bundled repo `skills/` and `agents/` (`bundle-assets` / root `postinstall`).
+- Remove cwd-based harness path resolution; consume bundled assets only.

package/README.md

@@ -0,0 +1,62 @@
+# @mstar-harness/opencode
+
+Morning Star (启明星) harness plugin for [OpenCode](https://opencode.ai).
+
+Install this package via OpenCode’s `plugin` array — it bundles `mstar-*` skills, role agents, and iteration commands so multi-role workflows (PM routing, SDD implement, QC tri-review, iteration lifecycle) work the same way as in the Cursor and Codex plugins.
+
+## Install
+
+Add to `opencode.json` (global or project):
+
+```json
+{
+  "plugin": ["@mstar-harness/opencode@latest"]
+}
+```
+
+Restart OpenCode.
+
+Or use the installer CLI:
+
+```bash
+npx @mstar-harness/cli init --target opencode
+```
+
+## What you get
+
+| Path in package | Contents |
+|-----------------|----------|
+| `harness-skills/` | `mstar-harness-core`, `mstar-iteration`, `mstar-sdd`, roles, phase/dispatch gates, … |
+| `harness-agents/` | Role shells (`project-manager`, `fullstack-dev`, `qc-specialist`, …) |
+| `harness-commands/` | `/iteration-start`, `/iteration-drive`, `/mstar-bootstrap` |
+
+The plugin resolves **only paths inside this package** — not `process.cwd()/skills`, so your app repo root does not affect harness loading.
+
+## Quick start
+
+1. Install the plugin (above).
+2. In OpenCode, start with the **Project Manager** agent (`project-manager`).
+3. For a full iteration: run **`/iteration-start`**, then **`/iteration-drive`** when plans are ready.
+
+Entry skill: **`mstar-harness-core`** (loaded before other `mstar-*` skills).
+
+## Docs
+
+- [INSTALL.md](./INSTALL.md) — setup, monorepo checkout, migration from legacy git plugin, troubleshooting
+- [Monorepo README](https://github.com/btspoony/mstar-harness#readme) — cross-host overview
+- [CHANGELOG.md](./CHANGELOG.md) — package release notes
+
+## Development (this monorepo)
+
+From the repository root:
+
+```bash
+bun install          # postinstall bundles harness-skills/ + harness-agents/
+bun run opencode:bundle-assets   # if you used --ignore-scripts
+```
+
+Plugin entry: `packages/opencode/src/mstar.ts` → `dist/mstar.js`.
+
+## License
+
+MIT — see [LICENSE](https://github.com/btspoony/mstar-harness/blob/main/LICENSE).

package/harness-commands/iteration-drive.md

@@ -30,6 +30,7 @@ Phase 2: Autonomous Execute  →  Phase 3: iteration-close  →  Phase 4: Create
 | 禁止（PM 线程） | 必须 |
 |-----------------|------|
 | Write/Edit/Shell 产品代码、写测试、跑 QC 审查（Phase 2） | 每条 implement/QC/QA Assignment ⇒ **1 次 `Task`** |
+| **多 task plan 用 inline 大包派发**（整份 plan / T1–Tn 贴进一个 dev Assignment） | **SDD**：`mstar-sdd` per-task 循环 — `task-brief` → implementer → `review-package` → task reviewer → `progress.md` |
 | 只写 Assignment 就进入下一 gate | 同轮 dispatch：`Subagent invokes issued: N`（N = Assignment 条数） |
 | 最后一个 plan `Done` 后直接开 PR / 汇报结束 | **Phase 3 → 4 → 5** 顺序执行 |
 | Phase 4 开 PR 后停止 | Phase 5 loop 至 merge-ready；**禁止**未过 §5.5 就结束会话 |
@@ -50,11 +51,12 @@ Phase 2: Autonomous Execute  →  Phase 3: iteration-close  →  Phase 4: Create
 
 1. `mstar-harness-core`
 2. `mstar-roles` → `references/project-manager.md`
-3. `skills/pm/SKILL.md` → **§ Host entry** + **§ Boot**（PM role identity + dispatch-first rules）
-4. `mstar-iteration` → **§ Phase 2–5**（§3 close gate；§4 PR delivery；§5 merge-ready loop）
-5. `mstar-compound` — Phase 3 §3.2 前加载（含 Phase 6 索引登记）
-6. `mstar-dispatch-gates` + host reference
-7. `mstar-plan-artifacts`, `mstar-plan-conventions`, `mstar-branch-worktree`
+3. `mstar-iteration` → **§ Phase 2–5**（§3 close gate；§4 PR delivery；§5 merge-ready loop）
+4. `mstar-compound` — Phase 3 §3.2 前加载（含 Phase 6 索引登记）
+5. `mstar-dispatch-gates` + host reference
+6. **`mstar-sdd`** — **before first implement dispatch** in Phase 2（per-task loop SSOT；多 task plan 默认 `Execution mode: sdd`）
+7. `mstar-review-qc` — before first QC dispatch in Phase 2
+8. `mstar-plan-artifacts`, `mstar-plan-conventions`, `mstar-branch-worktree`
 
 ## Phase 2: Autonomous Execute
 
@@ -74,9 +76,17 @@ Execute **`mstar-iteration` § Phase 2** exactly. Summary:
 4. **Integration branch** (§ 2.3) — `git checkout -b <spec_integration_branch> <iteration_base_branch>` when creating（**not** implicit `main`）
 5. **Per-plan loop** (§ 2.4) — for each non-`Done` plan:
    - Create plan feature branch from integration
-   - Dispatch implement subagents (dispatch-first)
+   - **SDD implement**（默认 `Execution mode: sdd`；`mstar-sdd` 已载入）— per task **串行**：
+     1. `sdd-workspace <plan-id>` → `{SDD_DIR}`
+     2. `task-brief <plan> N` → `{SDD_DIR}/task-N-brief.md`
+     3. Dispatch **one** implementer subagent（brief + report 路径；`references/implementer-prompt.md` 或 sticky 时 `implementer-continuation-prompt.md`）
+     4. `review-package BASE HEAD` → task diff
+     5. Dispatch **one** task reviewer subagent
+     6. Append `{SDD_DIR}/progress.md`；更新 `status.json` `task_commits[]`（如使用）
+     7. Next task — **never** parallel implementers；同 plan 同 dev 可用 **`SDD implementer session: sticky`**（`mstar-sdd/references/sticky-implementer-session.md`）
+   - **禁止**：把整份 plan 或 T1–Tn 全文贴进 implement dispatch；文件交接 SSOT → `mstar-sdd/references/file-handoffs.md`
    - Update `status.json` + main plan after each Completion Report v2
-   - QC **full tri-review** (`QC mode: full tri-review`, **N=3**, `mstar-review-qc`) + QA per plan
+   - QC **full tri-review** (`QC mode: full tri-review`, **N=3**, `mstar-review-qc`) + branch `review-package` + QA per plan
    - Merge plan branch → integration branch
    - Cross-plan progress sync → compass
    - Next plan
@@ -97,7 +107,7 @@ Execute **`mstar-iteration` § Phase 2** exactly. Summary:
 |------|---------|------|
 | 1 | §3.0 / §3.0.5 | Phase 边界；legacy compass 规范化（frontmatter + `## Roadmap Position`） |
 | 2 | §3.1 | 打印 close entry checklist，全部 `[x]` |
-| 3 | §3.2 | Compound；新增 doc 完成 `mstar-compound` Phase 6（`{KNOWLEDGE_DIR}/README.md`） |
+| 3 | §3.2 | Compound + **`<iteration-id>/` workspace promotion**；新增 doc 完成 `mstar-compound` Phase 6（`{KNOWLEDGE_DIR}/README.md`） |
 | 4 | §3.3 | `## Roadmap Position` current iteration → `delivered`；STRATEGY/tracker 按需更新 |
 | 5 | §3.4 | frontmatter `status: completed` + `end_date`；Compound Summary + Retrospective |
 | 6 | §3.5 | 打印 close exit checklist → commit → push |

package/harness-commands/iteration-start.md

@@ -1,6 +1,6 @@
 ---
 name: iteration-start
-description: Start a new harness iteration — research backlog, lock direction with grill-me, produce compass/plans, run mandatory Review & Edit chain via sequential Task dispatch (product-manager → architect → writing-specialist, each review-and-edit, then PM final lock), then create the integration branch from an explicit base. Not Done until review chain completes, compass is locked, and base/target branches are recorded.
+description: Start a new harness iteration — research, grill-me, compass/plans, Review & Edit chain (long-lived {SPECS_DIR}/ + {ITERATION_DIR}/<id>/ workspace; compound promotes workspace at close only), PM lock, integration branch.
 agent: project-manager
 ---
 
@@ -28,12 +28,11 @@ Detailed workflow → **`mstar-iteration` § Phase 1: iteration-start**（含 §
 
 1. `mstar-harness-core`
 2. `mstar-roles` → `references/project-manager.md`
-3. `skills/pm/SKILL.md` → **§ Host entry** + **§ Boot**
-4. `mstar-iteration` → **§ Phase 1: iteration-start**（迭代范围、compass 模板、§1.6 Review & Edit chain、状态初始化）
-5. `mstar-dispatch-gates`
-6. `mstar-phase-gates` → Prepare（specify → clarify → plan）
-7. `mstar-plan-conventions`, `mstar-plan-artifacts`
-8. `mstar-host` → active host reference（invoke 能力）
+3. `mstar-iteration` → **§ Phase 1: iteration-start**（迭代范围、compass 模板、**§1.5.5 产物边界**、§1.6 Review & Edit chain、状态初始化）
+4. `mstar-dispatch-gates`
+5. `mstar-phase-gates` → Prepare（specify → clarify → plan）
+6. `mstar-plan-conventions`, `mstar-plan-artifacts`
+7. `mstar-host` → active host reference（invoke 能力）
 
 ## 1. Research
 
@@ -99,12 +98,12 @@ Each role below **reviews and directly edits** the documents. Do not just flag i
 
 | # | Role | Required action | Gate |
 |---|------|-----------------|------|
-| 5.1 | **product-manager** | invoke: **edit** compass, plans, `{SPECS_DIR}/`; scope, UX, priorities | 完成后方可 5.2 |
-| 5.2 | **architect** | invoke: **edit** compass, plans, specs（含 5.1 修订后版本）; contracts, SQL, module boundaries | 5.1 返回后；完成后方可 5.3 |
-| 5.3 | **writing-specialist** | invoke: **edit** all iteration docs（含 5.1–5.2 修订后版本）; terminology, structure, clarity | 5.2 返回后 |
+| 5.1 | **product-manager** | invoke: **edit** compass, plans, **`{SPECS_DIR}/`**, **`{ITERATION_DIR}/<iteration-id>/`**（`guides/`、`specs/` 按需）— **禁止** `{KNOWLEDGE_DIR}/` 新增 | 完成后方可 5.2 |
+| 5.2 | **architect** | invoke: **edit** compass, plans, **`{SPECS_DIR}/`**, workspace `specs/`（含 5.1 后版本）— **禁止** `{KNOWLEDGE_DIR}/` 新增 | 5.1 返回后；完成后方可 5.3 |
+| 5.3 | **writing-specialist** | invoke: **edit** compass / plans / specs / iteration 文档; **specs corpus hygiene**（全库 `{SPECS_DIR}/` + 既有 `{KNOWLEDGE_DIR}/` 卫生与错放纠正，**不**新增 knowledge）— `mstar-iteration/references/iteration-artifact-boundaries.md` + `iteration-corpus-hygiene.md` | 5.2 返回后 |
 | 5.4 | **project-manager** | Merge subagent edits; resolve conflicts; **lock** compass (`status: locked`); confirm Prepare gates | 5.3 返回后 |
 
-**Evidence of done** = edited compass / plans / specs on disk + compass `status: locked`. **No** separate iteration review reports under `reports/` — unlike per-plan QC, there is no downstream audit chain to preserve.
+**Evidence of done** = edited compass / plans / **specs** / iteration docs on disk + specs corpus hygiene（及既有 knowledge 归档/错放纠正，如有）+ compass `status: locked`. **No** new `{KNOWLEDGE_DIR}/` from this chain — knowledge → **`mstar-compound`** @ iteration-close. **No** separate iteration review reports under `reports/`.
 
 **Tool rule (hosts with invoke)** — per **`mstar-dispatch-gates`** · specialist review-and-edit dispatch（**顺序链**，非 parallel batch）：
 
@@ -129,9 +128,9 @@ PM must print this block before §6; all `[ ]` must be `[x]`:
 
 - [ ] grill-me decisions recorded in compass
 - [ ] Draft compass + plans + `status.json` registered
-- [ ] product-manager invoke completed — compass / plans / specs edited
-- [ ] architect invoke completed — compass / specs edited
-- [ ] writing-specialist invoke completed — iteration docs edited
+- [ ] product-manager invoke completed — compass / plans / specs / **`<iteration-id>/` workspace**（按需）；**未**向 `{KNOWLEDGE_DIR}/` 新增
+- [ ] architect invoke completed — compass / **specs** edited；**未**向 `{KNOWLEDGE_DIR}/` 新增
+- [ ] writing-specialist invoke completed — specs + iteration docs edited；**specs corpus hygiene** done（全库 `{SPECS_DIR}/`；既有 knowledge 仅卫生/归档；错放已迁回 `iterations/` 或 archive）
 - [ ] PM final lock: compass `status: locked`; Prepare gates pass (blocked plans documented)
 - [ ] Branch policy locked: `iteration_base_branch`, `spec_integration_branch`, and `target_branch` recorded in compass / `status.json`
 - [ ] **THEN**: git commit + push `iteration/<iteration-id>`

package/harness-commands/mstar-bootstrap.md

@@ -14,11 +14,10 @@ Distill a coherent knowledge baseline from the current project — useful when t
 
 1. `mstar-harness-core`
 2. `mstar-roles` → `references/project-manager.md`
-3. `skills/pm/SKILL.md` → **§ Host entry** + **§ Boot**
-4. `mstar-plan-conventions`（路径符号）
-5. `mstar-strategy` → **§ STRATEGY.md structure** + **§ Creating STRATEGY.md**
-6. `mstar-compound` → **references/concepts-vocabulary.md**（CONCEPTS.md 规则）
-7. `mstar-compound-refresh` → **§ Core rules**（知识维护基线）
+3. `mstar-plan-conventions`（路径符号）
+4. `mstar-strategy` → **§ STRATEGY.md structure** + **§ Creating STRATEGY.md**
+5. `mstar-compound` → **references/concepts-vocabulary.md**（CONCEPTS.md 规则）
+6. `mstar-compound-refresh` → **§ Core rules**（知识维护基线）
 
 ## Phase 1: Survey — understand what exists
 

package/harness-skills/mstar-compound/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: mstar-compound
-description: Morning Star 知识结晶 —— 将已解决问题的经验沉淀为结构化知识文档，存入 `{KNOWLEDGE_DIR}`。支持 Bug track（症状/失败尝试/解决方案）与 Knowledge track（上下文/指南/适用场景）双轨。含重叠检测（更新已有文档）、可发现性检查（AGENTS.md 索引）、CONCEPTS.md 领域词汇协同。触发：迭代收口时（`mstar-iteration` § iteration-close）批量执行；也可独立触发。产出：`{KNOWLEDGE_DIR}/<category>/<slug>.md` + 可选 AGENTS.md 更新。
+description: Morning Star 知识结晶 —— 将已解决问题的经验沉淀为结构化知识文档，存入 `{KNOWLEDGE_DIR}`；迭代收口时从 `{ITERATION_DIR}/<iteration-id>/` 工作区提升 specs/guides。支持 Bug track 与 Knowledge track 双轨。含重叠检测、可发现性检查、CONCEPTS.md 协同。触发：iteration-close（`mstar-iteration` §3.2）或独立触发。产出：`{KNOWLEDGE_DIR}/<category>/<slug>.md`。
 ---
 
 # mstar-compound（知识结晶）
@@ -79,6 +79,26 @@ iteration-start → [plan lifecycle × N: specify→...→Done] → iteration-cl
 
 迭代内所有 plan Done 后，PM 回顾整轮迭代中产生的可结晶知识，批量 compound。per-plan Done 是 per-plan 的闭环终点；compound 是迭代级收口活动。
 
+### Iteration workspace promotion（iteration-close 强制盘点）
+
+正式迭代收口时，compound **除** plan 实现/debug/review 素材外，**必须**盘点当前迭代的 workspace：
+
+**路径**：`{ITERATION_DIR}/<iteration-id>/**`（含 `guides/`、`specs/`、扁平 `.md`；不含根目录 `*-delivery-compass.md` 除非 PM 显式纳入）。
+
+| 步骤 | 动作 |
+|------|------|
+| 1. Inventory | 列出 workspace 下全部 `.md`；读各文件 + workspace `README.md`（若有） |
+| 2. Triage | 每篇：**Promote** / **Keep snapshot** / **Skip**（理由写入 compound 摘要） |
+| 3. Promote | 值得跨迭代复用 → 走 Q1–Q8（或轻量判定）→ Phase 2 重叠检测 → Phase 3–6 **结构化重写**进 `{KNOWLEDGE_DIR}/`（**禁止**无改写整文件复制） |
+| 4. Trace | 源文件顶栏或 workspace README：`Promoted to: <knowledge-path>`；`{KNOWLEDGE_DIR}/README.md` 的 Source 可记 `iteration:<iteration-id>/<relpath>` |
+| 5. Summary | PM 写入 compass `## Compound Round Summary`：提升篇数、保留快照、跳过及原因 |
+
+**Promote 典型**：迭代 spec 已验证且指导未来实现；guide 含非显而易见的过程知识或失败尝试。
+
+**Keep snapshot**：仅迭代史、已被 `{SPECS_DIR}/` 取代的草案、或自检 ≤2 Yes 的琐碎笔记。
+
+边界 SSOT → **`mstar-iteration/references/iteration-artifact-boundaries.md`**。
+
 ## When to use (trigger)
 
 - **迭代收口时**（`mstar-iteration` § iteration-close）：PM 批量回顾所有 plan 的产物
@@ -114,7 +134,9 @@ In Cursor, Full mode dispatches subagents via Task tool. PM selects mode.
 
 ## Phase 1: Gather context
 
-Read the conversation history to understand:
+Read the conversation history **and**, when `iteration_id` is known, scan **`{ITERATION_DIR}/<iteration-id>/`** workspace per **Iteration workspace promotion** above.
+
+Understand:
 - What problem was solved (the concrete issue)
 - What was tried and didn't work
 - What the working solution was

package/harness-skills/mstar-compound-refresh/SKILL.md

@@ -45,7 +45,7 @@ For each candidate document, classify into one of five outcomes:
 3. **Match docs to reality.** When code differs from doc, update the doc — not ask whether the code change was "intentional."
 4. **Be decisive.** When evidence is clear (file renamed, class moved), apply. Only ask PM when genuinely ambiguous.
 5. **Avoid low-value churn.** Don't edit for typos, polish, or cosmetic changes that don't improve accuracy.
-6. **Delete, don't archive.** No `_archived/` directory. Git history is the archive. `git log --diff-filter=D -- <path>` finds deleted docs.
+6. **Delete, don't archive** — **except** formal **iteration-start** §1.6 corpus hygiene (`mstar-iteration/references/iteration-corpus-hygiene.md`), which **moves** superseded/redundant knowledge/specs to `{HARNESS_DIR}/archived/knowledge|specs/`. Outside that gate, git history is the archive; `git log --diff-filter=D -- <path>` finds deleted docs.
 7. **Evaluate document-set design.** Check whether two+ docs overlap and should be consolidated. Redundant docs silently drift apart.
 
 ## Scope selection

package/harness-skills/mstar-dispatch-gates/SKILL.md

@@ -69,7 +69,8 @@ description: Morning Star 派发与委派门禁 —— 仅 PM 可增派 subagent
 
 When **`Execution mode: sdd`** (`mstar-sdd`):
 
-- **串行**：one implementer at a time; one task reviewer after each — **never** parallel implementers (write conflicts).
+- **串行**：one implementer at a time; one **fresh** task reviewer after each — **never** parallel implementers (write conflicts).
+- **`SDD implementer session: sticky`**：same implementer subagent may **resume** across tasks when host supports it; **reviewers never resume** — see **`mstar-sdd/references/sticky-implementer-session.md`**.
 - File handoffs only — no pasted plan/diff/history in dispatch prompts.
 - Record per-task BASE SHA; use `review-package` for diffs — **never `HEAD~1`**.
 - After all tasks: branch `review-package` → **mandatory tri-review N=3** when `Execution mode: sdd`; **N=1** only for `inline` / explicit single override.
@@ -86,7 +87,7 @@ When **`Execution mode: sdd`** (`mstar-sdd`):
 
 - PM 写初稿；各角色通过宿主 invoke **直接编辑**目标文件（**不**另写仅评论式 `reports/` 替代修订）。
 - **1 Assignment ⇒ 1 invoke**。
-- **Phase 1 Review & Edit chain**（`mstar-iteration` §1.6）：**顺序** `product-manager` → `architect` → `writing-specialist`；上一角色落盘修订完成后再派发下一角色。**禁止**本链并行 batch（与 QC 三审不同）。
+- **Phase 1 Review & Edit chain**（`mstar-iteration` §1.6）：主产出 **`{SPECS_DIR}/`** + **`{ITERATION_DIR}/<iteration-id>/`** workspace；**禁止** start 链向 `{KNOWLEDGE_DIR}/` 新增。close 时 **`mstar-compound`** 提升 workspace → knowledge。
 - 其他彼此独立、无先后依赖的文档编辑任务：可并行（同条消息发满 N），见 **`parallel-dispatch.md`**。
 - PM 线程代做全部专业编辑 = **反模式**（`mstar-iteration` §1.6、`mstar-harness-core` 反模式索引）。
 - PM merge / lock（如 compass `status: locked`）在链末 subagent 返回后于 PM 线程完成。**不得**在 review-and-edit 链完成前 commit integration 分支。

package/harness-skills/mstar-host/references/codex.md

@@ -11,7 +11,7 @@ Parallel PM dispatch: read **`parallel-dispatch.md`** only when Codex exposes an
 - Plugin source: `.codex-plugin/plugin.json`.
 - Runtime skills: repo `skills/` mounted by the Codex plugin (`"skills": "./skills/"`).
 - Custom agent source: repo `codex/agents/*.toml`; CLI/manual install links these into `~/.codex/agents/` or project `.codex/agents/`.
-- `/pm`: shared force entry via the `pm` skill; after that, role behavior comes from `mstar-roles`.
+- **`/pm`** or **`pm` skill**: force PM entry → `mstar-roles` → `project-manager.md` (Codex primary; Cursor/OpenCode for general per-plan work). **`commands/`** when running iteration Phase 1–5.
 - Role files under root `agents/` are for hosts that load OpenCode/Cursor-style agent shells; Codex uses `codex/agents/*.toml` and still loads `mstar-roles` references directly.
 - Tool and plugin availability can be lazy-loaded or session-dependent. Use the tools actually present in the current session; do not infer capability from documentation alone.
 

package/harness-skills/mstar-host/references/cursor.md

@@ -6,7 +6,7 @@ Parallel PM dispatch: **`parallel-dispatch.md`** (Task tool uses same turn model
 
 ## Cursor-only context
 
-- Role prompts: `mstar-roles`; **`/pm`** → `project-manager` via `pm` skill + `mstar-roles`.
+- Role prompts: `mstar-roles`; **`/pm`** or **`pm` skill** → general PM orchestration (per-plan dispatch, gates, QC) without an iteration command. **`commands/`** (`iteration-start`, `iteration-drive`, `mstar-bootstrap`) only when running formal iteration Phase 1–5.
 - Routing-eval: `.cursor/skills/mstar-routing-eval/` — regression tooling only; not runtime load order.
 
 ## Plan mode × harness dual-write
@@ -44,6 +44,22 @@ Enforcement: `rules/mstar-cursor-plan-mode.mdc` when plugin active.
 - **`inline`**: **N=1** per `parallel-dispatch.md`.
 - SDD implement/reviewer: **serial** — see **`mstar-sdd`**.
 
+## SDD sticky implementer (Cursor Task resume)
+
+When Assignment has **`SDD implementer session: sticky`** (`mstar-sdd/references/sticky-implementer-session.md`):
+
+| Turn | Task tool |
+|------|-----------|
+| **First task** (or after `fresh` reset) | `subagent_type: <Execute as>` + `implementer-prompt.md` body; capture returned **agent id** → `{SDD_DIR}/implementer-session.json` → `host_agent_id` |
+| **Task 2+** | `resume: <host_agent_id>` + `implementer-continuation-prompt.md` body (new brief path + report path only) |
+
+**Rules:**
+
+- **1 implement dispatch ⇒ 1 Task** per task id (resume counts as the implement dispatch for that task).
+- **Task reviewer**: always **new** Task — **no** `resume` for reviewers.
+- After each task reviewer passes, PM updates `progress.md` and `implementer-session.json` `last_task`.
+- If `resume` fails or agent id missing → reset to **`fresh`** for that task; log reason in Status Update.
+
 ## Dispatch execution（canonical）
 
 Cursor PM dispatch = **`Task`** with `subagent_type` matching the Assignment `Execute as` role.

package/harness-skills/mstar-host/references/opencode.md

@@ -6,6 +6,7 @@ Parallel PM dispatch: **`parallel-dispatch.md`** (read in dispatch rounds).
 
 ## Role loading
 
+- **PM entry**: **`/pm`** or **`pm` skill** → `project-manager` for general orchestration; **`commands/`** (`iteration-start`, `iteration-drive`, `mstar-bootstrap`) for formal iteration only.
 - **Role shell**: `agents/<id>.md` referenced by `opencode.json` `agent.<id>` (frontmatter + role binding only).
 - **Role body**: `mstar-roles` `references/<id>.md` (or shared references + parameters).
 - Implementation evidence and RCA behavior: `mstar-coding-behavior`.
@@ -30,6 +31,8 @@ Harness **dispatch** on OpenCode = **one or more `task` tool calls**, each with
 
 PM workflow: finalize Assignment → **call task tool** with **subagent** + generated prompt → wait for subagent Completion Report → update plan / status.
 
+**SDD sticky implementer:** if the task tool exposes **resume** / agent id, follow **`mstar-sdd/references/sticky-implementer-session.md`** and the active host reference. If resume is **not** available, use **micro-batch** (2–3 tasks, one invoke) or **`SDD implementer session: fresh`** per task — do not assume sticky without host support.
+
 ## Role-mention hygiene (OpenCode)
 
 OpenCode may **auto-append** system lines when prompt text stacks multiple agent-id **prefix mentions**. Typical boilerplate (host-generated — **not** harness Assignment):

package/harness-skills/mstar-host/references/parallel-dispatch.md

@@ -63,4 +63,4 @@ Formal iteration Phase 2 uses the same SDD + tri rule — not a separate carve-o
 2. Prerequisite-only message? → **zero** batch dispatches unless `N = 1`.
 3. Dispatch message contains **exactly `N`** invocation calls?
 4. QC initial: **`Execution mode: sdd`** → **N=3**? **`inline`** → **N=1**? Targeted re-review → **N** = Assignment reviewer count?
-5. SDD implement → **serial** (never batch implementers)?
+5. SDD implement → **serial** (never batch implementers); sticky = **resume** same implementer, not parallel