claudedwithlove
explore/agent-docs

agent-docs

Cherished

Auto-discover and index library documentation for AI coding agents

·0··submitted April 15, 2026
View on GitHub
Clauded With Love Rating
8.1 / 10

agent-docs automatically discovers and indexes library documentation from multiple sources (llms.txt, MCP, Context7) to create a static reference for AI coding agents. It solves the documented problem of agents failing to reliably discover documentation by creating a persistent index that's always in context.

Code Quality6.5
Usefulness8.7
Claude Usage7.2
Documentation9.1
Originality8.9
Highlights
  • Addresses a real, measured problem with concrete data (56% failure rate for auto-discovery)
  • Multi-source documentation discovery with intelligent fallbacks and caching
  • Excellent README with clear problem statement, solution explanation, and comprehensive usage examples
To Improve
  • Add TypeScript type definitions and interfaces to improve code maintainability
  • Include test coverage and error handling examples for edge cases like network failures or malformed package.json files