Intent-Based Code Review

Review code changes by verifying that acceptance criteria are met, rather than inspecting every line of implementation. This approach is designed for AI-generated code at scale, where the volume and velocity of changes exceed human cognitive capacity for line-by-line review. Detailed code inspection is triggered only when acceptance criteria fail, focusing human attention where it matters most.

When to Use This Skill

Use this skill when:

• Reviewing AI-generated code where acceptance criteria are well-defined
• PR volume exceeds the team's capacity for line-by-line review
• The change was produced by a structured workflow (e.g., research-plan-implement) with a REQUEST.md or specification
• You want to shift review effort from "Did the agent write it correctly?" to "Does it meet the requirements?"
• You are reviewing changes produced by a multi-model orchestration workflow

Do NOT use this skill when: