docmole

Docmole is an MCP server that integrates with AI assistants like Claude and Cursor to enable natural language search across documentation sites. It supports two modes: Local RAG Mode, which crawls documentation and generates embeddings with OpenAI and LanceDB for self-hosted search, and Mintlify Mode, which proxies requests to Mintlify's AI Assistant API for zero-setup access. The tool works with any documentation site, performs hybrid semantic and keyword matching, and maintains multi-turn conversation context. Operations teams and developers benefit from instant access to large documentation volumes without manual searching.

[{"step":"Identify your documentation source. Ensure Docmole is configured to access the relevant documentation (e.g., AWS Docs, internal Confluence pages, or GitHub Wiki). Use tools like Cursor or Claude with the Docmole plugin to enable seamless integration.","tip":"If you're using internal documentation, upload it to Docmole’s knowledge base or ensure it’s accessible via a public URL. For APIs like AWS, Docmole typically has pre-loaded documentation."},{"step":"Formulate a natural language query. Instead of searching for 'AWS Lambda cold start,' try 'How can I reduce latency in my serverless API?' This yields more context-rich results.","tip":"Include context in your query, such as your use case (e.g., 'for a high-traffic e-commerce platform') or specific pain points (e.g., 'causing timeout errors')."},{"step":"Review the summarized results. Docmole will provide key points, links, and references. Focus on the 'Actionable Next Steps' section for immediate tasks.","tip":"Cross-reference the sources to validate the information. Docmole prioritizes official documentation, but always double-check critical details."},{"step":"Iterate if needed. If the results aren’t specific enough, refine your query (e.g., 'How does Provisioned Concurrency work in AWS Lambda?') or ask for additional details on a particular technique.","tip":"Use follow-up queries like 'Show me code examples for implementing X' or 'What are the cost implications of Y?' to dive deeper."},{"step":"Integrate the findings into your workflow. Apply the recommendations directly (e.g., updating Lambda configurations) or share the results with your team for collaboration.","tip":"For team use, export the Docmole results as a markdown file or share the links directly in your project management tool (e.g., Jira, Notion)."}]