AgentSkillsCN

breakdown-epic

将宏大叙事分解为可落地的用户故事，并明确验收标准、估算工时、依赖关系以及冲刺周期的分组安排。此技能适用于宏大叙事的规划与待办事项清单的日常维护。

SKILL.md
--- frontmatter
name: breakdown-epic
description: Break down epics into implementable user stories with acceptance criteria, estimates, dependencies, and sprint groupings. This skill should be used for epic planning and backlog grooming.
acceptance:
  - components_identified: "Major components and sub-features identified from epic"
  - stories_created: "User stories created in proper format with acceptance criteria"
  - estimates_provided: "Story points estimated using complexity/effort/risk"
  - dependencies_mapped: "Dependencies between stories identified and documented"
  - sprints_suggested: "Sprint groupings suggested based on dependencies and priorities"
inputs:
  epic_description:
    type: string
    required: true
    description: "Epic or feature description to break down"
  tech_stack:
    type: array
    required: false
    description: "Technology stack for technical considerations"
outputs:
  story_files:
    type: array
    description: "List of generated story file paths"
  epic_file:
    type: string
    description: "Path to generated epic summary file"
  total_points:
    type: number
    description: "Total story points across all stories"
  story_count:
    type: number
    description: "Number of stories created"
  sprint_count:
    type: number
    description: "Suggested number of sprints"
telemetry:
  emit: "skill.breakdown-epic.completed"
  track:
    - story_count
    - total_points
    - sprint_count
    - priority_distribution
    - dependency_count
    - duration_ms

Breakdown Epic Skill

Purpose

Transform high-level epics or features into detailed, actionable user stories ready for sprint planning and implementation.

Core Capabilities:

  • •Epic scope analysis and component identification
  • •User story creation (As a... I want... So that...)
  • •Story point estimation (complexity/effort/risk)
  • •Dependency mapping between stories
  • •Sprint grouping suggestions
  • •Story and epic file generation

Prerequisites

  • •Configuration file (.claude/config.yaml) with planning settings
  • •Epic description (20+ words minimum)
  • •Story point scale defined (default: Fibonacci 1,2,3,5,8,13,21)

Workflow

Step 0: Load Configuration and Epic Description

Action: Load configuration and parse epic requirements.

Load config: Read .claude/config.yaml for planning settings (storiesLocation, epicsLocation, defaultVelocity, storyPointScale, techStack)

Get epic description: From user input (20+ words minimum), parse scope, business goals, technical requirements

Prepare directories: Create stories and epics directories

See: references/templates.md#configuration-format and #step-0-configuration-loading for details

Halt if:

  • •Configuration missing
  • •Epic description <20 words (too vague)
  • •Cannot create directories

See: references/epic-analysis-guide.md for scope analysis

Step 1: Analyze Epic Scope

Action: Break down epic into major components and sub-features.

Identify components:

  1. •Core Features - What functionality is needed?
  2. •User Types - Who will use this?
  3. •Technical Requirements - What tech/integrations?
  4. •Security Requirements - What needs protection?
  5. •Non-Functional Requirements - Performance, scalability, etc.

Output: Epic analysis with major components, sub-features, user types, technical/security requirements, NFRs

See: references/templates.md#step-1-epic-analysis-example for complete analysis format

Group by themes:

  • •Related features together
  • •Similar technical complexity
  • •Common user journeys

See: references/epic-analysis-guide.md for detailed analysis techniques

Step 2: Create User Stories

Action: Convert components into user stories with acceptance criteria.

Story format: "As a [user], I want [capability], So that [benefit]" with acceptance criteria, priority, component

See: references/templates.md#step-2-user-story-format for format and 5 complete examples

Story quality checks:

  • •Clear user perspective
  • •Specific, testable acceptance criteria (2-5 per story)
  • •Appropriate priority
  • •Properly scoped (not too large, not too small)

See: references/story-creation-guide.md for story writing best practices

Step 3: Estimate Story Points

Action: Estimate each story using complexity, effort, and risk.

Estimation factors: Complexity (1-5), Effort (1-5), Risk (1-3)

Calculate: Story Points = Complexity + Effort + Risk, round to Fibonacci (1, 2, 3, 5, 8, 13, 21)

See: references/templates.md#step-3-estimation-factors-template for detailed scales and examples

See: references/estimation-guide.md for detailed estimation techniques

Step 4: Identify Dependencies

Action: Map dependencies between stories.

Dependency types:

1. Blocks: Story A must complete before Story B can start

code
story-auth-001 (Signup) BLOCKS story-auth-003 (Email Verification)

2. Related: Stories share components but can be done in parallel

code
story-auth-001 (Signup) RELATED story-auth-002 (Login)

Dependency rules:

  • •Foundation before features (models before APIs)
  • •Authentication before authorization
  • •Core functionality before enhancements
  • •Infrastructure before applications

Output: Dependency graph with BLOCKS/REQUIRES/RELATED relationships, critical path identification

See: references/templates.md#step-4-dependency-examples for complete graph format

See: references/dependency-mapping-guide.md for dependency analysis

Step 5: Suggest Sprint Groupings

Action: Group stories into sprints based on dependencies, priorities, and velocity.

Grouping rules:

  1. •Priority first - P0 stories in early sprints
  2. •Dependencies respected - Blocked stories after blockers
  3. •Velocity constraints - Don't exceed team capacity
  4. •Logical grouping - Related stories together

Output: Sprint groupings with story assignments, velocity constraints, dependency respect, logical grouping

See: references/templates.md#step-5-sprint-grouping-example for complete sprint plan format

Step 6: Generate Story Files and Epic Summary

Action: Create story markdown files and epic summary.

Generate files: Use bmad-commands write_file for each story and epic summary

Output: Story files (complete format with user story, ACs, dependencies, estimation, DoD) and epic summary file

See: references/templates.md#step-6-story-file-format-template and #step-6-epic-summary-file-format for complete structures

Step 7: Present Summary to User

Action: Present epic breakdown summary for review.

Output: Breakdown complete summary with story count, points, sprints, files, priorities, sprint plan, dependencies

See: references/templates.md#step-7-summary-output-template for complete format

Output

Return structured JSON with story_files, epic_file, metrics (total_points, story_count, sprint_count), and telemetry

See: references/templates.md#json-output-format for complete structure

Best Practices

  1. •Epic Scope - Keep epics focused (10-15 stories max)
  2. •Story Size - Target 3-8 points per story (not too large)
  3. •Clear ACs - 2-5 specific, testable acceptance criteria
  4. •Priority Discipline - Not everything can be P0
  5. •Dependency Mapping - Identify blockers early
  6. •Velocity Reality - Don't overcommit sprint capacity

See: references/story-creation-guide.md for best practices

Routing Guidance

Use this skill when:

  • •Starting new epic or major feature
  • •During backlog grooming sessions
  • •Planning multiple sprints ahead
  • •Breaking down product requirements

Always use before:

  • •sprint-plan skill (sprint planning)
  • •refine-story skill (detailed story refinement)

Feeds into:

  • •create-task-spec skill (for implementation specs)

Reference Files

Detailed documentation in references/:

  • •templates.md: All output formats, config examples, story/epic file templates
  • •epic-analysis-guide.md: Epic scope analysis, component identification
  • •story-creation-guide.md: User story format, acceptance criteria, best practices
  • •estimation-guide.md: Story point estimation techniques
  • •dependency-mapping-guide.md: Dependency types, mapping, sprint grouping

Part of BMAD Enhanced Planning Suite