AgentSkillsCN

run-epic

通过完整的 BMAD 流程（创建故事 → 开发故事 → 代码评审 → 端到端测试 → 提交 PR），自主执行史诗中剩余的所有故事。

SKILL.md
--- frontmatter
name: run-epic
description: Autonomously execute all remaining stories in an Epic through the full BMAD pipeline (create-story → dev-story → code-review → e2e-tests → PR)

Run Epic - Autonomous Story Pipeline

Execute the complete BMAD development pipeline for all remaining stories in the specified Epic.

Prerequisites

  • •Epic file exists in _bmad-output/planning-artifacts/
  • •Git repository initialized with remote
  • •GitHub CLI (gh) authenticated

Pipeline Per Story

  1. •Branch: Create feature branch from main
  2. •SM Agent: Create story via bmad-sm subagent
  3. •Dev Agent: Implement story via bmad-dev subagent
  4. •Review Agent: Code review + auto-fix via bmad-reviewer subagent
  5. •QA Agent: E2E tests (if frontend) via bmad-qa subagent
  6. •Git: Commit, push, create PR, merge

Usage

Provide the Epic identifier or file path:

code
Run the epic pipeline for Epic-001

Subagent Invocation Pattern

Each phase spawns a specialized subagent:

code
use_subagent with agent_name="bmad-{role}" and autonomous instructions

Risk Acknowledgment

This runs autonomously without confirmations. Ensure:

  • •Tests are comprehensive
  • •CI/CD will catch issues
  • •You can revert if needed