IT IS CRITICAL THAT YOU FOLLOW THESE STEPS - while staying in character as the current agent persona you may have loaded:
<steps CRITICAL="TRUE"> 1. Always LOAD the FULL @{project-root}/_bmad/core/tasks/workflow.xml 2. READ its entire contents - this is the CORE OS for EXECUTING the specific workflow-config @{project-root}/_bmad/bmm/workflows/4-implementation/code-review/workflow.yaml 3. Pass the yaml path @{project-root}/_bmad/bmm/workflows/4-implementation/code-review/workflow.yaml as 'workflow-config' parameter to the workflow.xml instructions 4. Follow workflow.xml instructions EXACTLY as written to process and follow the specific workflow config and its instructions 5. Save outputs after EACH section when generating any documents from templates </steps>code-review
开展对抗式的高级开发者代码评审，在每条故事中精准定位3至10个具体问题。从代码质量、测试覆盖率、架构合规性、安全性、性能等多个维度发起挑战，绝不满足于“看起来不错”——必须找出最低限度的问题，并在获得用户同意后，支持自动修复。
SKILL.md
--- frontmattername: 'code-review' description: 'Perform an ADVERSARIAL Senior Developer code review that finds 3-10 specific problems in every story. Challenges everything: code quality, test coverage, architecture compliance, security, performance. NEVER accepts `looks good` - must find minimum issues and can auto-fix with user approval.'