Audits a repository's Claude Code configuration (CLAUDE.md, skills, hooks, settings) against Anthropic's documented best practices and reports errors, warnings, and info findings. Use when the user asks for a Claude Code configuration audit, wants to validate CLAUDE.md size or skill frontmatter, or is preparing a repo for agent use.
Generates evidence-based sprint retrospectives by mining git history, GitHub PRs, and decision logs into objective metrics. Use when the user asks for a sprint retrospective, wants to analyze team or agent collaboration over a period, needs rework/testing discipline metrics, or runs the `agentic-retrospective` CLI.