Both Context7 and docmancer solve the same problem: giving AI coding agents access to up-to-date library documentation. They take fundamentally different approaches.
Feature Comparison
| Feature | docmancer | Context7 |
|---|---|---|
| Delivery | Local CLI (PyPI) | Cloud API (MCP server) |
| Index location | Local SQLite FTS5 on your machine | Remote server |
| Offline support | Full offline after docmancer docs add | Requires network |
| API keys | None for the core path | Required for higher tiers |
| Agent support | Claude Code, Cursor, Cline, Codex, Gemini, GitHub Copilot, OpenCode, Claude Desktop | MCP-compatible agents |
| Privacy | All indexing and queries stay local | Queries sent to remote API |
| Project config | docmancer.yaml (local-first, reproducible) | Per-query configuration |
| Agent memory | Discovers and shares what your agents already wrote | Not covered |
| Pricing | Free MIT package; optional paid sync and team workspaces | Free tier with limits |
| Open source | Yes (MIT-licensed PyPI package) | Partial |
When to choose docmancer
Choose Docmancer if you want documentation indexed and searched on your own machine, and if you also want the decisions, preferences, and rules your coding agents already recorded to be readable and shared between them. Docs retrieval is one half of the product; the other half is Shared Memory, which Context7 does not attempt.
When to choose Context7
Choose Context7 if you prefer a zero-install cloud API approach and your workflow does not require offline access or reproducible local indexes.
Try docmancer
pipx install docmancer --python python3.13
docmancer setup
docmancer docs add https://react.dev
docmancer docs query "how do I use useEffect cleanup?"