Morning Star Harness

Plugin morning-star-harness from marketplace mstar-local in btspoony/mstar-harness. Commands read from the repo's manifest — today.

Skills (27)

• mstar-routing-eval

  [Cursor maint] Morning Star 路由与 prompt 迭代评估 —— 维护/回归专用（非日常 PM 编排 runtime）。路径：本仓库 `.cursor/skills/mstar-routing-eval/`。(1) PM 路由回归：`assets/routing-evals.json`；(2) Prompt/规则迭代评估；(3) Routing Eval Report 模板。变更 PM 路由或 harness 门禁、`@prompt-engineer` 改 agents/skills、或显式跑场景集时 Read。

• grill-me

  Interview the user relentlessly about a plan or design until shared understanding, resolving each decision branch. Use when a plan or design needs stress-testing, or the user says 'grill me'.

• mstar-artifacts

  Plan artifacts: main plans, durable review summaries, QC/PM Acceptance bundles, workflow state/config/progress/evidence, issue capture, residual severity. Read when authoring or mapping them.

• mstar-audit

  Audit or survey a codebase into an evidence-backed, prioritized review report (read-only): bugs, security, perf, test gaps, tech debt, DX, a PR/branch/diff (`pr` variant), test-suite audit.

• mstar-branch-worktree

  Branch authorization, control/integration/source checkout isolation, L1/L2 writable-task isolation, QC and PM Acceptance checkout alignment, guarded worktree/branch cleanup after merge.

• mstar-coding-behavior

  Use for non-trivial implementation, debugging, refactoring or code review in any project: requirement drift, speculative complexity, agent-facing interfaces. Not workflow or role selection.

• mstar-conventions

  harness 路径约定：`{HARNESS_DIR}`/`{PLAN_DIR}`/`{SDD_DIR}`/`{ITERATION_DIR}`/`{KNOWLEDGE_DIR}`/`{SPECS_DIR}`/`{ADR_DIR}`/`{REVIEW_DIR}` 等符号的发现与初始化、docs 与 harness 子树边界、Spec 分支模型。读写 `.mstar/` 或开 plan 前必读。

• mstar-design-md

  DESIGN.md 设计系统规范：创建/审计/维护 design tokens、三级完整性清单、light/dark 双主题（DESIGN.dark.md）。用户提到 DESIGN.md / design tokens / design system，或实现、审查 UI 时读取。

• mstar-design-system

  Use when aligning work to an existing baseline (prototype, design export, product markup, or design contract) and classifying each difference as defect, decision, or phantom. Not for greenfield design or `DESIGN.md` authoring (`mstar-design-md`).

• mstar-dispatch-gates

  Dispatch gates: PM-only leaf dispatch, executor/delegation identity, anti-recursion, parallel readiness, SDD QC tri-review, inline single-seat exceptions. Read before dispatch or leaf execution.

• mstar-doc-corpus

  Use for writing, auditing or maintaining the three-layer document corpus — `{KNOWLEDGE_DIR}` engineering knowledge, `{SPECS_DIR}` current-state specs and `{ADR_DIR}` decision records — covering layer boundaries, section skeletons, naming, lifecycle and the optional validator. No harness lifecycle or

• mstar-e2e

  Separately requested E2E / real-browser / device / installed-deployment verification. Load only for an explicit request or `amazing-e2e-check`; never from routine development acceptance or missing screenshots.

• mstar-engine-legacy

  Engine-absent execution-state boundary: without a CLI/engine, execution state is conversation tracking (no-plan mode) and no file execution route is recreated. Use only when the host has no engine capability.

• mstar-harness-core

  Lifecycle and authorization authority: delivery loop, state machine, Done ownership, task-category routing, topic index. Use for lifecycle, authorization, Done, routing, or CLI/plugin version drift.

• mstar-host

  Host adapter for OpenCode, Cursor, Codex, Kimi, ZCode, omp, dsh: entry, dispatch, plan mode, parallel dispatch, skill/command loading. Detect from session tools, then read `references/<host>.md`.

• mstar-iteration

  Start, drive, resume or close an iteration (also without a slash command): Phase 1 direction lock, autonomous execute, iteration-close, PR delivery, merge-ready loop, post-merge close.

• mstar-knowledge

  Load for `{KNOWLEDGE_DIR}` capture, maintain/refresh/bootstrap, or close-time promotion（结晶/清理/显式 bootstrap/收口提升）；skip with reasoned skipped when nothing to crystallize. Not for ordinary feature coding.

• mstar-phase-gates

  Spec-Driven 双阶段门禁：Prepare（specify→clarify→plan）、Execute（plan(locked)→tasks→implement，多 task 默认 SDD）；意图门禁、clarify 纪律、hotfix 压缩路径、可验证编辑、Phase Gate 证据。

• mstar-pr-loop

  Use only when the user explicitly starts a review→fix loop on an existing ready PR, or Phase 1 records `pr_merge_policy: reviewed-auto`. Never under `reviewed-bot`, and never for routine one-shot review requests.

• mstar-project-governance

  项目治理：roadmap 内容在 `{HARNESS_DIR}/store.db` 的权威读写（import/export、revision-guarded replace）、milestone 结构化目标与 issue 关联、issue capture、`_default` 归属。

• mstar-prose

  Use when prose looks like a leaked reasoning transcript (session/PR vantage, change narration, hedges, planning residue) and needs a HEAD-resolvability check. Not for skill structure/triggers (`mstar-skill-authoring`) or coding behavior.

• mstar-review-qc

  QC orchestration: SDD mandatory plan QC tri-review (`qc1`…`qc3` + consolidated), inline/hotfix single seat, tri identity gate, residual registration. Read before dispatching or consolidating QC.

• mstar-roles

  Role prompt hub and single load-selection authority: identity-first `references/<role>.md`, PM-activated skill presets, the none/named-preset/resume decision. Always load for any Morning Star role.

• mstar-sdd

  Subagent-driven development (SDD): file handoff, per-task implementer + reviewer (L2), progress ledger, review package for plan QC tri (L3), fresh or sticky session. Read under Execution mode: sdd.

• mstar-skill-authoring

  Use when creating/rewriting a skill (`SKILL.md`), narrowing a description trigger, compressing a body, reviewing skill structure, or editing model-visible non-body text (schemas, CLI help, refusals, Assignment fields). Not for app features.

• mstar-use-cli

  Use when choosing, running or interpreting `mstar`/`mstar-harness` CLI commands: family selection, semantic preconditions (control root, cwd, identity, revision/token), exit codes and refusal codes.

• pm

  PM entry shim: force `project-manager` orchestration when the user invokes `/pm` or this skill. Boot, routing and dispatch SSOT live in `mstar-roles/references/project-manager.md`.

Slash commands (7)

• /morning-star-harness:amazing-e2e-check

  Run explicitly requested E2E, browser, device, or installed-deployment scenarios in an independent verification workflow.

• /morning-star-harness:amazing-pr-review

  Deep pre-merge review of a PR / branch / diff at quick / default / deep strength: one tally-derived verdict, findings, and the posted GitHub review when a PR number exists. Read-only advisory — never approves, requests changes, or merges. Not for self-reviewing your own change.

• /morning-star-harness:amazing-test-audit

  Read-only test-suite audit — junk patterns, value/retention bars, prioritized delete/repair/consolidate findings, or a whole-subsystem campaign. Use to prune or repair a test suite; not for writing tests, product fixes, or routine acceptance.

• /morning-star-harness:codebase-audit

  Read-only codebase survey — a prioritized, evidence-backed review report of what is worth doing. Use standalone or before /iteration-start; scope keywords pick the effort level and focus. Never edits source.

• /morning-star-harness:iteration-drive

  Drive or resume the selected locked iteration through Phase 2→6 — autonomous execute → iteration-close → PR → merge-ready loop → post-merge close. Takes no arguments.

• /morning-star-harness:iteration-loop

  Run a whole iteration Phase 1→6 autonomously — code-first direction lock with a retained prototype, then execute → close → PR → merge-ready → post-merge close. Optional direction hint and scale S|M|L|XL; explicit autonomous opt-in, no grill-me, no routine confirmation.

• /morning-star-harness:iteration-start

  Start a new harness iteration — research, interactive direction lock, retained design prototype and confirmation, compass/plans, selected specialist Review & Edit, PM lock, integration branch — then auto-continue Phase 2→6 unless `pause` is given.

Subagents (12)

• architect

  Architect - system design, technical decisions, and technical documentation (architecture notes, ADRs, interface contracts).

• code-reviewer

  Code Reviewer - SDD per-task review (L2) + codebase audit execution (audit). Read-only seat: does not implement, fix, or occupy a QC seat.

• frontend-dev

  Frontend Developer - pages/components/interactions/accessibility/frontend performance. This is the default frontend owner in fullstack work (split with `@fullstack-dev`) and the preferred role for pure UI tasks.

• fullstack-dev-2

  Fullstack Developer (Track 2) - the second implementation track parallel to `@fullstack-dev` (independent modules/APIs/page islands). PM should assign this role when tasks can run in parallel or when acceleration is needed, with explicit module boundaries and branch ownership.

• fullstack-dev

  Fullstack Developer - backend-led fullstack implementation (API/business/data). PM should split UI-heavy or new-page work to `@frontend-dev`, and use `@fullstack-dev-2` for the second parallel track. One developer can handle hotfixes or small single-stream updates.

• ops-engineer

  Ops Engineer - deployment, monitoring, and infrastructure operations, including CI/CD and observability.

• product-manager

  Product Manager - requirements analysis, product planning, market/user research, and product-facing documentation.

• prompt-engineer

  Prompt Engineer - design and optimize prompts and skills for agents, including refactoring and debugging prompt systems.

• qc-specialist-2

  Quality Control Specialist (Reviewer #2) - code review (diff, logic, security/correctness). Not a test runner.

• qc-specialist-3

  Quality Control Specialist (Reviewer #3) - code review (diff, logic, performance/reliability). Not a test runner.

• qc-specialist

  Quality Control Specialist (Reviewer #1) - code review (diff, logic, maintainability). Not a test runner.

• writing-specialist

  Writing Specialist - documentation, fiction, copywriting, and script writing.

Hooks (3)

• SessionStart

  node

• PreToolUse

  node

• PreToolUse

  node

Similar plugins