跳到正文
cccskills
免费在 GitHub 公开

task-planning

Generates phased, dependency-ordered implementation tasks from specifications. Use after spec is complete and before starting implementation.

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(4)

  • SKILL.md4.1 KB
  • modules/dependency-patterns.md6.0 KB
  • modules/phase-structure.md4.9 KB
  • modules/tech-stack-patterns.md2.3 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Task Planning

Overview

Transforms specifications and implementation plans into concrete, dependency-ordered tasks. Creates phased breakdowns that guide systematic implementation.

When To Use

  • Converting specifications to implementation tasks
  • Planning feature implementation order
  • Identifying parallel execution opportunities
  • Breaking down complex features into phases

When NOT To Use

  • Writing specifications - use spec-writing

Task Phases

Tasks follow a 5-phase structure from setup through polish:

  • Phase 0: Setup - Project initialization, dependencies, configuration
  • Phase 1: Foundation - Data models, interfaces, test infrastructure
  • Phase 2: Core Implementation - Business logic, APIs, services
  • Phase 3: Integration - External services, middleware, logging
  • Phase 4: Polish - Optimization, documentation, final testing

For detailed phase definitions, selection guidelines, and anti-patterns, see modules/phase-structure.md.

Task Format

Each task includes:

  • ID: Unique identifier (TASK-001)
  • Description: Clear action statement
  • Phase: Which phase it belongs to
  • Dependencies: Tasks that must complete first
  • Parallel Marker: [P] if can run concurrently
  • Files: Affected file paths
  • Criteria: How to verify completion

Dependency Rules

Dependencies define execution order and identify parallelization opportunities:

  • Sequential Tasks: Execute in strict order when dependencies exist
  • Parallel Tasks [P]: Can run concurrently when ALL nonconflicting conditions are met
  • File Coordination: Tasks affecting same files MUST run sequentially

Nonconflicting Criteria for Parallel Execution:

  • ✅ Files: No file overlap between tasks
  • ✅ State: No shared configuration or global state
  • ✅ Dependencies: All prerequisites satisfied
  • ✅ Code paths: No merge conflicts possible
  • ✅ Outputs: Tasks don't need each other's results

Mark tasks with [P] ONLY if they pass ALL criteria above.

For fan-out/fan-in patterns, task ID conventions, and validation rules, see modules/dependency-patterns.md.

Example Task Entry

## Phase 2: Core Implementation

### TASK-007 - Implement user authentication service [P]
**Dependencies**: TASK-003, TASK-004
**Files**: src/services/auth.ts, src/types/user.ts
**Criteria**: All auth tests pass, tokens are valid JWT

Verification: Run pytest -v to verify tests pass.

Quality Checklist

  • All requirements mapped to tasks
  • Dependencies are explicit
  • Parallel opportunities identified
  • Tasks are right-sized (not too large/small)
  • Each task has clear completion criteria

Related Skills

  • spec-writing: Creating source specifications
  • speckit-orchestrator: Workflow coordination

Exit Criteria

  • Every task has a unique ID in TASK-NNN format and includes Description, Phase (0-4), Dependencies list, Parallel marker ([P] or sequential), Files, and Criteria fields
  • Tasks marked [P] pass all 5 nonconflicting criteria: no file overlap, no shared state, dependencies satisfied, no merge conflicts possible, no output dependency on each other
  • Every functional requirement from the source spec is mapped to at least one task; unmapped requirements are flagged
  • Quality checklist passes all items before output is finalized
  • Tasks within each phase are ordered so that no task depends on one with a higher task number within the same phase

评价

还没有评价。欢迎分享你的使用感受。

评论

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。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Shapes turns action-first: action leads, steps numbered, state restated. Use for ADHD-friendly output. Do not use to trim tokens; use response-compression.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

Inverts burden of proof for code additions. Use when reviewing PRs, planning refactors, or running unbloat to challenge every addition's necessity.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

Tracks per-agent token usage and flags waste in parallel dispatch. Use when evaluating parallel agent efficiency or after a multi-agent run.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

Coordinates Claude agent teams via filesystem protocol. Use when orchestrating parallel agents with task dependencies. Do not use for single-agent tasks.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

Evaluates API surface design, consistency, and exemplar alignment. Use when reviewing public API changes or before releasing a new API surface.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

Selects architecture paradigm via research before scaffolding. Use when architecture is undecided and the choice needs justification and documentation.

中文简介正在准备中。当前显示原文说明。

athola/claude-night-market3432026年10月17日 更新

查看 athola 的全部技能

报告此技能的问题