The sprint completeness oracle — a mechanical, exit-code-driven check that every feature in a sprint manifest is provably complete across three dimensions (a done-record reference, resolvable edge tests, and a resolvable wiring test). Fails loud with a per-feature gap list naming the missing dimension. Read-only. Use to gate a sprint before it is called done (Slice A of the sprint-mode epic), or any time you want a completeness reading.
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
sprint-oracle reads a sprint manifest (authored by sprint-plan), scans the test tree for the @sprint-feature / @kind tags, and reports whether every feature is complete. It is the answer to "the final system ships incomplete": a sprint is not done until the oracle exits 0. Read-only — it reports, it never edits.
A feature is complete only when all three hold:
done_record is non-empty (the feature traces to a spec AC).edge_tests resolves to a test tagged // @sprint-feature:<id> @kind:edge.wiring_test resolves to a test tagged // @sprint-feature:<id> @kind:wiring."Resolves" means: the named test exists in a scanned .mjs file and its tag's feature id and kind match. A test tagged @kind:happy never satisfies an edge or wiring requirement — the happy path does not count as completion.
| Code | Meaning |
|---|---|
0 | every feature complete (or the manifest has zero features — vacuously complete) |
2 | gaps found — at least one feature is missing a dimension; a per-feature gap list is printed to stderr |
1 | operational error — the manifest could not be read or parsed (distinct from a completeness gap) |
A gap is {feature, dimension, detail} where dimension is one of done-record / edge / wiring, and detail names the unresolved test(s).
node .claude/skills/sprint-oracle/oracle.mjs <manifest-path> [test-root]
test-root defaults to tests. The CLI prints each gap as GAP <feature>: <dimension> — <detail> and exits with the code above. For programmatic use, import runOracle({manifestPath, testRoot}) → {code, gaps, error?}.
Example gap output:
GAP search: edge — no resolvable @kind:edge test among [search_empty_query]
GAP search: wiring — wiring_test 'search_end_to_end' does not resolve to a @kind:wiring test
No script can prove a test is semantically an edge case or a true end-to-end exercise. The oracle verifies that the manifest's claims resolve to tagged tests — it trades unbounded gameability for mechanical checkability. That is a deliberate, documented bound (see the locked Q3 decision in the mvp-sprint-parallel-cycles spec): far better than today's no-record-of-done, and crisp enough to grep.
sprint-plan (which authors the manifest + validates its shape). Keep authoring and gating separate.还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Phase 10.5 — move the slug's workflow artifacts (intake, scout, research, spec, approvals, swarm state, security reports, rendered diagrams) to docs/archive/<YYYY-MM-DD>/<slug>/. Runs before /commit so the committed tree is clean of work-in-flight files. workflow.json stays live and gets archived as the first step of /commit.
中文简介正在准备中。当前显示原文说明。
Drift check between the baseline implementation on disk and the claims in `docs/init/seed.md` + cross-references in CLAUDE.md, README.md, and the rendered docs site. Verifies hook/agent/skill/command names + counts, settings.json wiring, project.json key presence, .mcp.json servers, vendored license files, and helper script presence. Exit 0 PASS / 1 FAIL — suitable for CI. Read-only; safe to invoke any time.
中文简介正在准备中。当前显示原文说明。
PM-mode brainstorm helper. Captures the requirement via Socratic dialogue before any entry phase (`/intake`, `/spec`, `/tdd`) drafts its artifact. Stage 0 skip-check, Stage 1 gap-analysis, Stage 2 probe-loop, Stage 3 confirm-and-persist. Output lives at `docs/brief/<slug>.md`. Never proposes solutions — Stage 2 dialogue discipline is structurally enforced via `discipline.mjs`.
中文简介正在准备中。当前显示原文说明。
Draft a Business Requirements Document (BRD) for cross-functional or stakeholder-heavy work that needs more structure than an intake. Use after `/intake` when the request spans multiple systems/teams, carries regulatory weight, or needs formal sign-off. Output lives at `docs/brd/<slug>.md`.
中文简介正在准备中。当前显示原文说明。
Workflow track for tasks that need no TDD — documentation edits, governance count bumps, vendored-skill content updates, configuration tweaks, formatting, typo fixes, dependency bumps where no project code changes. Skips `/scenario` and `/implement` (no failing test to drive) and runs the work directly. `archive`, `memory-sync`, `/grant-commit`, and `/commit` remain mandatory. `verify`, `simplify`, `integrate`, and `document` are conditional — required when the diff hits one of the listed triggers, optional otherwise. `verify` is skipped only when the diff is pure-docs/prose AND `project.json → test.kind` is `behavior` (absent/invalid `test.kind` → `structural` → verify runs). Chore is a stripped-down pipeline, not a bypass; never silently skip a conditional phase whose triggers apply.
中文简介正在准备中。当前显示原文说明。
Analyze a codebase and recommend Claude Code automations (hooks, subagents, skills, plugins, MCP servers). Use when user asks for automation recommendations, wants to optimize their Claude Code setup, mentions improving Claude Code workflows, asks how to first set up Claude Code for a project, or wants to know what Claude Code features they should use.
中文简介正在准备中。当前显示原文说明。