Search and read your WhatsApp messages with a read-only MCP server on Cloudflare Workers
An open-source project deploys a small server to your own Cloudflare account and links it to your phone as a WhatsApp linked device. Any MCP client can then search and read your chats, but cannot send messages.
The problem
WhatsApp has no official API for reading your own personal messages, so AI agents cannot query your chat history. People who want to ask an agent about past conversations have no straightforward way to give it access. Contact names are also often missing, which leaves many chats identified only by phone number.
Steps
- Clone the repo, run npm install, and create a Cloudflare KV namespace for OAuth tokens, then paste its id into wrangler.jsonc.
- Set an OWNER_PASSWORD secret of 12 or more characters and deploy with npm run deploy on a Workers Paid plan.
- Open the deployed Worker URL and link WhatsApp by scanning the QR code under Linked Devices, or use the phone number option; the first link imports history.
- Add the server to your agent, for example with claude mcp add --transport http whatsapp pointing at your /mcp URL, then sign in with the password via /mcp.
- Optionally ask the agent to find names for unnamed chats; it looks up numbers in a contacts tool it already has and saves them with set_contact_names.
From the official docs
claude mcp add --transport http whatsapp https://your-url/mcpResults
- The agent gets read-only tools to list chats, read a chat, run full-text search across all messages, and trigger a sync.
- Messages are stored in a SQLite database inside a Durable Object in your own Cloudflare account, and the server pulls new messages on request and once a day.
- Limitations noted by the author: it is unofficial and against WhatsApp's terms, history import depends on what WhatsApp hands over, media is shown only as labels, and one deployment serves one owner.
As reported by the source (GitHub repository README (finereli/whatsapp-mcp)); AgentGid did not measure these figures.
This suits a solo developer who is comfortable with npm, Wrangler and secrets, and who is fine paying for a Workers Paid plan. Keep in mind that it is unofficial and against WhatsApp's terms, so the linked account carries some risk. Claude Code is $20/mo, while OpenCode (Free + $10/mo) or Pi (Free (OSS)) can act as the MCP client if you want to avoid that cost.
The agent used here
Similar use cases
Run a self-hosted life dashboard that Claude can control through an MCP server
Edris Ranjbar's open-source Edi Life OS combines habits, goals, Kanban, finance, focus timers and notes in one self-hosted PHP and MySQL dashboard. An optional MCP server lets Cla…
The MCP server exposes 15 tools covering the dashboard, goals, tasks, habits, finances and notes.
Let Claude Code and Codex point at Mac screen elements with bigarrow
An open-source, MIT-licensed macOS command-line tool called bigarrow, with a skill for Claude Code and Codex, lets AI agents draw a click-through arrow and text sign over the scre…
Drawing needs no macOS permission; only some target-finding modes need Accessibility or Screen Recording, granted to the terminal app that launched b…
Route between coding-agent models with the open-source Weave Router 2.0
The Weave team presented Weave Router 2.0 in a Show HN post. It is an open-source routing model that sits in front of coding agents and switches between LLMs per turn, and the aut…
As reported by the Weave author on Hacker News: on Terminal Bench 4.0 and SWE Atlas, the router had equivalent pass rates to GPT-6 Astra.
Guides
Claude Code Deploys: 6 Open-Source Repos for a Containment System, Checked
A viral post says GitHub Actions, Argo CD, Argo Rollouts, Prometheus, Sentry and OpenTelemetry turn Claude Code deploys into a containment system. We checked each: role, license, what needs Kubernetes, and what Claude Code's own permissions add.
How to Cut Token Usage in Claude Code and Cursor (2026)
Practical ways to spend fewer tokens with Claude Code and Cursor: measure first, keep context small, match the model to the task, filter what the agent reads, and the open-source tools that help.
Sonnet Builds, Opus Reviews, Haiku Reads: Does the Claude Code Advisor Save Money?
A viral post says running Sonnet, Haiku and Opus together in Claude Code gives better code for fewer tokens. We checked it against Anthropic's docs and prices: when the advisor pays off, when it costs more, and why a ready project kit matters more than the model mix.
JEV for Claude Code: A Second Opinion on Every Command, Checked
A viral post says jev-axi screens every command your Claude Code agent runs for a fraction of a cent. We read the repo, TypeSafe's docs and privacy policy: what runs locally, what is sent, how it fails, and how it compares with auto mode.