finereli/whatsapp-mcp is a read-only MCP server that lets AI agents search and read a user's WhatsApp conversations. MCP (Model Context Protocol) is the protocol that connects LLMs to tools and data. The server runs on Workers (a serverless runtime) in the user's own Cloudflare account. It is written in TypeScript, and the repository was created on October 18, 2026, with no public release version confirmed.
Once a WhatsApp account is connected, MCP clients can search and read that account's conversations. Messages are stored in the SQLite database of a Cloudflare Durable Object, and when a read request arrives and the stored content is older than 10 minutes, new messages are fetched from WhatsApp; it also syncs once a day. The distinguishing point of this repository is that serverless setup: it collects personal messages through a linked device and exposes them as read and search tools. The license is MIT and the repository has 313 GitHub stars.
That said, the server does not support sending messages, and it neither downloads media nor converts it to voice. Contact names can be stored. Deployment has to be prepared in the user's own Cloudflare account, and a paid Workers plan is required. The WhatsApp integration is unofficial, so a terms-of-service risk remains. The provided information holds no figures for judging search quality or performance. The quality score is 5.0 out of 10, but that value is a recorded evaluation score, not a performance measurement. If the plan is to use personal messages as a read-only tool, it is worth reviewing, but it does not fit use cases that expect sending features or search performance.