AgentSkillsCN

plan

将待开发的范围或功能集拆解为一个个故事与任务,以便开展冲刺规划(索菲娅·中村的工作流程)

SKILL.md
--- frontmatter
name: plan
description: Break a scope or feature set into stories and tasks for sprint planning (Sofia Nakamura's workflow)
disable-model-invocation: true

Plan the following scope: $ARGUMENTS

Step 1 — Understand the Scope

  • Read the request and identify all deliverables
  • Read the existing codebase to understand what exists and what's new
  • Identify which layers are affected (see Layers in project config)

Step 2 — Break Into Stories

Each story is a vertical slice of functionality that can be independently delivered and demonstrated.

Format each story as:

markdown
### Story: <title>
**As a** <user/developer/operator>
**I want** <capability>
**So that** <business value>

**Acceptance Criteria:**
- [ ] <specific, testable criterion>

**Files Affected:**
- <layer/file> — <what changes>
- <test file> — <what tests needed>

**Estimated Complexity:** Small / Medium / Large

Step 3 — Order by Dependency

  • Identify dependencies between stories (Story B requires Story A)
  • Order stories so dependencies are satisfied
  • Mark stories that can be parallelized

Step 4 — Break Stories Into Tasks

Each story should have 2-6 concrete tasks:

markdown
**Tasks:**
1. [ ] Add Pydantic models to the models file
2. [ ] Add service method to the service layer
3. [ ] Add API endpoint to the appropriate router
4. [ ] Write tests (success, validation, error paths)
5. [ ] Update project documentation and env example file
6. [ ] Run /verify and /review

Step 5 — Risk Assessment

For each story, identify:

  • Technical risks — unknown APIs, complex logic, concurrency concerns
  • Dependencies — external libraries, config changes, breaking changes
  • Testing complexity — needs mock implementations, special fixtures

Step 6 — Output Sprint Plan

markdown
# Sprint Plan: <scope title>

## Summary
- Stories: X
- Total tasks: X
- Estimated effort: Small/Medium/Large

## Story Order (dependency-aware)
1. Story A (no dependencies)
2. Story B (depends on A)

## Stories
<full story details from Step 2>

## Risks
<from Step 5>

Rules

  • Stories must be independently demonstrable (Sofia's requirement)
  • Every story must include tests (Priya's requirement)
  • Every story must maintain backwards compatibility (Sofia's requirement)
  • Architecture decisions documented with rationale (Alex's requirement)