Cosa-SAI (Gemini Docs)

Loads entire documentation sets directly into Gemini's 2M token context window, eliminating traditional RAG limitations for compre

m-gonzalo 14
Claude CodeClaude DesktopGeneric
View source ↗

Gemini Docs MCP Server enables AI assistants to access comprehensive documentation for various technologies using Google's Gemini API with its 2M token context window. Built by M-Gonzalo, it overcomes traditional RAG system limitations by eliminating the need for chunking, custom retrievers, and vector databases, instead loading entire documentation sets directly into the LLM. The server provides specialized tools for checking task feasibility, getting problem-solving hints, evaluating code practices, and exploring implementation alternatives. This approach delivers well-reasoned answers that consider entire technology specifications, making it particularly valuable for learning, debugging, and exploring alternative approaches across both common and obscure technologies.

Source

Repository: https://github.com/m-gonzalo/cosa-sai

Maintain Cosa-SAI (Gemini Docs)?

Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.

[Cosa-SAI (Gemini Docs) on getagentictools](https://getagentictools.com/mcp/m-gonzalo-cosa-sai?ref=badge)