What problem does it solve?
This Skill provides an end-to-end workflow to read a code repository and produce an evidence-based, structured review as Markdown plus a standalone HTML report, including architecture maps, entrypoint traces, and actionable recommendations.
Core Features & Use Cases
- Evidence-based reports: generate a comprehensive review of a repo's structure, core abstractions, entrypoints, and extension points.
- MD+HTML deliverables: produce a portable, offline-friendly report with Mermaid diagrams.
- Getting started: ideal for onboarding developers or conducting repository audits to communicate findings to stakeholders.
Quick Start
Use the repo-deep-dive workflow on a target repository to generate docs/repo_review.md and docs/repo_review.html. Example:
- Generate a repo snapshot: python3 agent/skills/repo-deep-dive-report/scripts/repo_snapshot.py --repo . --output docs/_repo_snapshot.md
- Create the review content in docs/repo_review.md following the outline in references/report_outline.md
- Render to HTML: python3 agent/skills/repo-deep-dive-report/scripts/render_md_to_html.py --input docs/repo_review.md --output docs/repo_review.html --title "Repo Review"