spec-first

Install, configure, verify, and refresh required harness runtime readiness facts for spec-first workflows on Claude Code, Codex, Kiro, Qoder, or Cursor.

公开 workflow：编写、改写、迁移或按 audit findings 修复 spec-first source skill（skills/<name>/）时使用。不要用于一次性回答、解释/总结/翻译、只审计、文档导出、第三方安装、普通 spec-* workflow 执行，或手改 generated runtime mirrors。

Use this standalone skill when the user asks to mine a repo's existing coding conventions for future AI coding, generate or refresh project rules with AGENTS.md/CLAUDE.md pointers, create Cursor or Qoder rule files from actual code evidence, or make AI-generated code follow a specific project's habits. Do not use for team standards governance, normal code review/debug/refactor work, linter/formatter configuration, generic best practices, unsupported tool rule files such as .cursorrules or .kiro/steering rules, or generated runtime mirror edits.

Build applications where agents are first-class citizens. Use this skill when designing autonomous agents, creating MCP tools, implementing self-modifying systems, or building apps where features are outcomes achieved by agents operating in a loop.

Capture a visual demo reel (GIF, terminal recording, screenshots) for PR descriptions. Use when shipping UI changes, CLI features, or any work with observable behavior that benefits from visual proof. Also use when asked to add a demo, record a GIF, screenshot a feature, show what changed visually, create a demo reel, capture evidence, add proof to a PR, or create a before/after comparison.

Create, share, view, comment on, edit, and run human-in-the-loop review loops over markdown documents via Proof, the collaborative markdown editor at proofeditor.ai ("Proof editor"). Use when the user wants to render or view a local markdown file in Proof, share markdown to get a URL, iterate collaboratively on a Proof doc, comment on or suggest edits in Proof, HITL a spec/plan/draft for human review, sync a Proof doc back to local, or work from a proofeditor.ai URL. Trigger on phrases like "view this in proof", "share to proof", "iterate with proof", or "HITL this doc", and on spec-brainstorm / spec-ideate / spec-plan handoffs for human review. Also match clear requests for a rendered/shared markdown review surface even if the user does not name Proof. Do not trigger on "proof" meaning evidence, math/logic proof, burden of proof, proof-of-concept, or bare "proofread this" requests where inline text review is expected.

Review code diffs, PRs, or branch implementation changes using tiered persona reviewers, confidence-gated findings, and merge/dedup synthesis. Do not use for requirements/plan/task-pack document review, implementation execution, planning unresolved work, or commit/push/PR creation.

Refresh existing learning or pattern docs under docs/solutions/ when they are stale, outdated, overlapping, drifted, inaccurate, or explicitly named for refresh/consolidation against the current codebase. Use for docs/solutions/ refresh/audit/sweep/cleanup/consolidation requests, specific stale solution docs, or a narrow stale-doc hint from spec-compound. Do not use for new learning capture, active debugging, general refactor/migration/code-review work, non-docs/solutions documentation sweeps, transcript archiving, mandatory completion gates, or generated runtime mirror edits.

Document a just-solved, source-confirmed problem as reusable team knowledge in docs/solutions/. Do not use for active debugging, unresolved hypotheses, one-off summaries or transcript archiving, mandatory completion gates, or refreshing existing learnings; use spec-compound-refresh for stale knowledge.

Systematically find root causes and fix bugs. Use when debugging errors, investigating test failures, reproducing bugs from issue trackers (GitHub, Linear, Jira), or when stuck on a problem after failed fix attempts. Also use when the user says 'debug this', 'why is this failing', 'fix this bug', 'trace this error', 'why is this slow', or reports a 'performance regression', or pastes stack traces, error messages, or issue references.

Review requirements, plans, task packs, or Markdown planning artifacts for coherence, feasibility, scope, risk, and downstream readiness. Do not use for code diff/PR/branch implementation review, implementation execution, or PR merge-readiness code review. When reviewer dispatch is available and explicitly authorized for the current host, run bounded parallel reviewers; when dispatch is unavailable, unauthorized, explicitly disabled, or unsafe, fall back to a single-agent report-only review instead of bypassing host boundaries.

Run metric-driven iterative optimization loops. Define a measurable goal, build measurement scaffolding, then run parallel experiments that try many approaches, measure each against hard gates and/or LLM-as-judge quality scores, keep improvements, and converge toward the best solution. Use when optimizing clustering quality, search relevance, build performance, prompt quality, or any measurable outcome that benefits from systematic experimentation. Inspired by Karpathy's autoresearch, generalized for multi-file code changes and non-ML domains.

Create a structured HOW plan when the user has a clear goal, requirements document, brainstorm/PRD handoff, feature, bug, or project; also deepen an existing plan when asked to strengthen the plan as a whole. Prefer spec-brainstorm for unresolved WHAT/product exploration, spec-work for implementation or tests, spec-doc-review for independent plan review, and spec-write-tasks for task-pack compilation. Do not use for generated runtime mirror fixes.

Public workflow entrypoint (spec-prd): create, write, refine, or validate planning-readiness of brownfield PRD-grade requirements for existing systems before implementation planning. Do not use for 0-1 product exploration, unresolved product shape, HOW planning/task compilation, implementation/debug/review, lightweight direct fixes, generated runtime mirror edits, or PRD/design-source/source consistency audits; route to spec-brainstorm, spec-plan/spec-write-tasks, spec-work, review workflows, or spec-app-consistency-audit as appropriate.

Search and ask questions about your coding agent session history. Use when asking what you worked on, what was tried before, how a problem was investigated across sessions, what happened recently, or any question about past agent sessions. Also use when the user references prior sessions, previous attempts, or past investigations — even without saying 'sessions' explicitly.

Audit agent skills for trigger precision, scope boundaries, input/output contracts, workflow quality, progressive disclosure, eval readiness, safety, maintainability, and runtime governance. In spec-first repositories, also audit source skills, dual-host governance, generated runtime drift, and alignment with spec-first workflow principles. Do not use for installing third-party skills, creating new skills, or general code review.

Govern source-backed team development standards: query confirmed rules, audit health, initialize or draft evidence-based candidates, and prepare promotion/deprecation proposals. Do not restore legacy standards workflow entrypoints, skills/spec-standards/, or treat advisory candidates as hard context.

Execute a settled plan, validated task pack, spec path, or concrete implementation request in the current repo. Do not use when WHAT/HOW is unresolved, target repo scope is ambiguous, task-pack freshness is stale/unverifiable, scope would expand beyond the plan, or the fix requires hand-editing generated runtime mirrors.

Public workflow entrypoint (spec-write-tasks): compile a settled local spec-plan into an optional derived task pack for spec-work, or validate an existing local task pack before execution. Use for explicit plan-splitting/task-doc requests or high-complexity work suitability; do not use for plan authoring, implementation execution, unresolved scope, small low-risk plans, progress/approval state, remote/generic task lists, or generated runtime mirror edits. Keep the plan as single source of truth; tasks are derived and optional.

Use before substantial work in a spec-first repo, and when users ask what spec-first workflow or command to run next. Route non-trivial edits, state-changing commands, debugging, review, planning, setup/update, optimization, and architecture/prompt/workflow/contract decisions. Do not use for lightweight factual answers, current-context explanations, narrow lookups, user-supplied single-document summaries, or clearly scoped low-risk small edits.

Discover WHAT for a selected feature/problem before PRD or planning when behavior, scope, users, success criteria, or handoff context remain unresolved. Do not use for open-ended ideation, brownfield PRD authoring/refinement/validation, clear HOW planning/task compilation, implementation/debug/review/setup, generated runtime mirror fixes, factual answers, or cleanup.

Audit mobile App PRD/Figma/local-source consistency across page routes, KMP/Clean Architecture, components, analytics, i18n, engineering quality, and industry lenses before runtime validation; use for cross-source App consistency, not ordinary code review, PRD authoring, build/test execution, UI polish, or product-code edits.

Internal helper for public workflows that need an isolated git worktree for parallel feature work or PR review; use only when delegated by `spec-work`, `spec-code-review`, or another public workflow.

Summarize recent spec-first releases, or answer a specific question about a past release with a version citation. Use when the user types `spec-release-notes` or asks "what changed in spec-first recently?" or "what happened to `<skill-name>`?".

Generate and critically evaluate grounded ideas about a topic. Use when asking what to improve, requesting idea generation, exploring surprising directions, or wanting the AI to proactively suggest strong options before brainstorming one in depth. Triggers on phrases like 'what should I improve', 'give me ideas', 'ideate on X', 'surprise me', 'what would you change', or any request for AI-generated suggestions rather than refining the user's own idea.

Commit, push, and open a PR with an adaptive, value-first description. Use when the user says "commit and PR", "push and open a PR", "ship this", "create a PR", "open a pull request", "commit push PR", or wants to go from working changes to an open pull request in one step. Also use when the user says "update the PR description", "refresh the PR description", "freshen the PR", "rewrite the PR body", "write a PR description", "draft a PR description", or "describe this PR" — the skill will produce a description without committing or pushing if that is all the user wants. Produces PR descriptions that scale in depth with the complexity of the change, avoiding cookie-cutter templates.

Create a git commit with a clear, value-communicating message. Use when the user says "commit", "commit this", "save my changes", "create a commit", or wants to commit staged or unstaged work. Produces well-structured commit messages that follow repo conventions when they exist, and defaults to conventional commit format otherwise.

Resolve PR review feedback by evaluating validity and fixing issues with conflict-aware resolver dispatch. Use when addressing PR review comments, resolving review threads, or fixing code review feedback.

Search Slack for interpreted organizational context -- decisions, constraints, and discussion arcs that shape the current task. Produces a research digest with cross-cutting analysis and research-value assessment, not raw message lists. Use when searching Slack for context during planning, brainstorming, or any task where organizational knowledge matters. Trigger phrases: 'search slack for', 'what did we discuss about', 'slack context for', 'organizational context about', 'what does the team think about', 'any slack discussions on'. Differs from slack:find-discussions which returns individual message results without synthesis.

[BETA] Start the dev server, open the feature in a browser, and iterate on improvements together.

This skill should be used when generating and editing images using the Gemini API (Nano Banana Pro). It applies when creating images from text prompts, editing existing images, applying style transfers, generating logos with text, creating stickers, product mockups, or any image generation/manipulation task. Supports text-to-image, image editing, multi-turn refinement, and composition from multiple reference images.

Run browser tests on pages affected by current PR or branch

Build web interfaces with genuine design quality, not AI slop. Use for any frontend work - landing pages, web apps, dashboards, admin panels, components, interactive experiences. Activates for both greenfield builds and modifications to existing applications. Detects existing design systems and respects them. Covers composition, typography, color, motion, and copy. Verifies results via screenshots before declaring done.

Create engaging changelogs for recent merges to main branch

Clean up local branches whose remote tracking branch is gone. Use when the user says "clean up branches", "delete gone branches", "prune local branches", "clean gone", or wants to remove stale local branches that no longer exist on the remote. Also handles removing associated worktrees for branches that have them.

Report a bug in the spec-first plugin

This skill should be used when writing Ruby and Rails code in DHH's distinctive 37signals style. It applies when writing Ruby code, Rails applications, creating models, controllers, or any Ruby file. Triggers on Ruby/Rails code generation, refactoring requests, code review, or when the user mentions DHH, 37signals, Basecamp, HEY, or Campfire style. Embodies REST purity, fat models, thin controllers, Current attributes, Hotwire patterns, and the "clarity over cleverness" philosophy.

Build and test iOS apps on simulator using XcodeBuildMCP. Use after making iOS code changes, before creating a PR, or when verifying app behavior and checking for crashes on simulator.