reporead_skill

Overview

This skill analyzes GitHub repositories using RepoRead AI to produce documentation, architecture breakdowns, security audits, diagrams, and LLM-optimized summaries. It integrates with an MCP server (preferred) or a REST API fallback and supports token-aware workflows for efficient usage. Use it to quickly understand, document, or vet any public GitHub repository archived on clawhub-style sources. The skill is implemented in Python and focuses on practical outputs you can act on.

How this skill works

The skill imports a repository, queues the requested analysis type, and polls lightweight status until the job completes. Analysis types include technical, readme, security, mermaid, and llmstxt, each returning structured results you can fetch and use as context. It works by calling MCP tools when available or equivalent REST endpoints, and it enforces token checks before starting analyses to avoid unexpected failures.

When to use it

• You need a fast orientation or architecture summary before making changes to a codebase.
• You want a full README or updated documentation generated from code and comments.
• You need a pre-deploy security audit or a PR-level vulnerability scan.
• You want visual system diagrams (Mermaid) to explain workflows and components.
• You are building LLM-driven tools that require repo-context in LLM-friendly text.

Best practices

• Check token balance (get_token_balance) before starting analyses to prevent interruptions.
• Use list_repositories to avoid importing the same repo twice.
• Poll get_analysis_status every ~10 seconds instead of fetching full results repeatedly.
• Run readme + mermaid in parallel to get docs and diagrams together.
• Inspect error fields on failed analyses; do not retry automatically without investigating.

Example use cases

• Onboard a new contributor by running a technical analysis to produce an architecture brief.
• Generate a production-ready README and diagrams for an archived clawhub skill repository.
• Run a security analysis before merging an external PR or deploying a forked repo.
• Produce an LLM-optimized summary (llmstxt) for downstream automation or agent context.
• Automate periodic backups by listing imported repos and re-running readme analyses for archive snapshots.