evernote-mcp-server

Evernote MCP Server is a local MCP integration that enables Claude Desktop and other MCP-compatible LLMs to search and retrieve your Evernote notes using natural language. It uses OAuth 1.0a authentication for secure access and stores tokens persistently for seamless re-authentication. The server supports read-only operations including note searches, retrieval, and content access, making it ideal for teams that need to incorporate Evernote data into AI-assisted workflows. Recent versions include Docker containerization with security-hardened Chainguard images, remote HTTP/JSON-RPC support, and production-ready stability improvements that eliminate restart cycles.

1. **Set up the Evernote MCP server**: Install the Evernote MCP server in your AI tool (e.g., Claude Desktop) and authenticate via OAuth to connect to your Evernote account. Ensure you have the correct API permissions. 2. **Define your query**: Use the prompt template to specify the tags, timeframes, or other filters you want to apply to your note retrieval. For example, replace [TAG] with `#project-alpha` and [TIMEFRAME] with `30 days`. 3. **Execute the query**: Run the prompt in your AI tool to fetch the relevant notes. The MCP server will retrieve the notes and pass them to the AI for analysis. 4. **Analyze and act**: Review the AI’s summary and suggestions. Use the extracted details (e.g., deadlines, action items) to prioritize tasks or draft follow-up communications. For example, copy the action items into your task manager or calendar. 5. **Refine as needed**: If the results aren’t specific enough, adjust the prompt (e.g., narrow the timeframe or add additional tags) and rerun the query. Use the MCP server’s filtering capabilities to refine your search.