菜单
Use this skill to break work into small, verifiable chunks. Activates when planning implementation, breaking down features, or when tasks seem too large.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | micro-tasking |
| description | Use this skill to break work into small, verifiable chunks. Activates when planning implementation, breaking down features, or when tasks seem too large. |
Break work into 2-5 minute tasks with clear completion criteria.
Large tasks fail. Small tasks succeed.
A task that takes more than 10 minutes is not a task—it's a project that needs decomposition.
Every micro-task has:
task:
id: TASK-001
title: Short, action-oriented description
duration: 2-5 minutes
file: exact/path/to/file.ts
line_hint: ~50-75 # approximate location
action: |
Specific action to take
verification: |
How to know it's done
depends_on: [] # or [TASK-000]
feature: "Add user login"
tasks:
- Create LoginForm component skeleton (3 min)
- Add email input with label (3 min)
- Add password input with label (3 min)
- Add submit button (2 min)
- Add form onSubmit handler (3 min)
- Add loading state to button (3 min)
- Add error message display (3 min)
- Connect to auth API (5 min)
- Add success redirect (3 min)
- Write unit test for form render (5 min)
- Write unit test for submission (5 min)
bug: "Login fails silently"
tasks:
- Add console.log to trace request (2 min)
- Verify API endpoint is correct (2 min)
- Check response handling (3 min)
- Add error state to component (3 min)
- Display error message to user (3 min)
- Add test for error case (5 min)
refactor: "Extract shared button styles"
tasks:
- Identify all button variants (3 min)
- Create Button component file (2 min)
- Add base button styles (3 min)
- Add variant prop (primary/secondary) (3 min)
- Replace first usage (3 min)
- Replace second usage (3 min)
- Verify styles match original (2 min)
- Run visual regression test (3 min)
- id: TASK-XXX
title: Create [Component] skeleton
duration: 3 min
file: src/components/[Component].tsx
action: Create component with props interface, return null
verification: Component imports without error
- id: TASK-XXX
title: Add test for [behavior]
duration: 5 min
file: src/components/[Component].test.tsx
action: Write describe block with it() for [behavior]
verification: Test runs (can be failing)
- id: TASK-XXX
title: Add error handling for [case]
duration: 3 min
file: src/lib/[file].ts
line_hint: ~45
action: Add try/catch around [operation]
verification: Error case shows message instead of crashing
For each task:
1. Announce: "Starting TASK-XXX: [title]"
2. Set timer: 5 minutes max
3. Do the ONE thing
4. Verify completion criteria
5. Commit if appropriate
6. Announce: "Completed TASK-XXX"
7. If over time: Task was too big, note for future
| Aspect | Large Tasks | Micro-Tasks |
|---|---|---|
| Progress visibility | Unclear | Clear |
| Context switching | Loses state | Easy resume |
| Estimation accuracy | Way off | Reasonable |
| Verification | Hard | Easy |
| Morale | Frustrating | Satisfying |
Works with:
/ledger - Track micro-task completion/execute-plan - Execute task by task/riper - Plan phase creates micro-tasksWHEN to auto-invoke: Creating API routes, building endpoints, adding route.ts files, implementing REST/GraphQL APIs, adding authentication to APIs, rate limiting, API validation with Zod, handling HTTP methods (GET/POST/PUT/DELETE). WHEN NOT to invoke: Pure frontend work, styling, React components without API calls, documentation-only tasks, database migrations without API changes.
WHEN to auto-invoke: Code review requests, refactoring code, cleaning up code, performance optimization, security audits, linting issues, reducing technical debt. WHEN NOT to invoke: Building new features (use domain-specific skills), initial implementation, documentation-only work.
WHEN to auto-invoke: End of productive sessions, after solving difficult problems, after user corrections, after discovering effective workarounds, after establishing project conventions. WHEN NOT to invoke: Trivial sessions, sessions with only exploration, when explicitly told not to learn.
WHEN to auto-invoke: Database schema design, creating migrations, writing SQL queries, query optimization, Supabase operations, Prisma/Drizzle schema changes, PostgreSQL tasks, RLS policies, indexes. WHEN NOT to invoke: Pure frontend work, API logic without database, styling, React components, deployment tasks.
WHEN to auto-invoke: Creating CI/CD pipelines, GitHub Actions workflows, Docker/containerization, Vercel/Netlify deployment, infrastructure as code, environment variables, deployment configurations. WHEN NOT to invoke: Application code changes, UI development, database schema design, testing application logic.
WHEN to auto-invoke: Measuring code quality, testing AI outputs, establishing success criteria, defining acceptance tests, checking implementation correctness. WHEN NOT to invoke: Simple implementations, prototyping, exploration phases, when speed matters more than measurement.