Why WireCat exists
A document in one chat, a customer's question in another, a promise you made last week. Finding the right message takes time, and it's easy to miss something that needs your attention.
Messengers came first: they hold most of what people agree on, and nothing gave an AI agent easy access to them. From there WireCat grew to the rest of your personal context: mail, meetings, documents, calendar and notes.
What WireCat is today
A set of command-line tools, one for each source. Each works with your own account, keeps a searchable copy on your computer and lets an AI agent read, search and prepare replies. WireCat Desktop installs, updates and checks them for you, signs you in and gives your agents the instructions to use them, with no terminal needed.
You choose the agent and what it may do. Any agent that runs terminal commands can use the tools: Claude Code, Codex, Cursor, Gemini CLI, Hermes and others. Apps without a terminal, such as ChatGPT or Claude, connect through MCP, a standard for linking AI apps to tools.
How it is built
- Your data stays on your computer. Messages, mail and documents are kept in a local store. WireCat has no servers to send them to; your agent receives only what a task needs.
- Your accounts, your limits. Every command names the profile it uses, and a profile can be read-only or limited to chosen chats. Nothing is sent, marked read or deleted unless you asked for it.
- Built for agents. One operation per call, one JSON result and a fixed exit code for each failure. Every tool ships an MCP server and a skill that teaches your agent its commands.
- Open source. Every tool and the desktop app are on GitHub under the Apache 2.0 licence. Read the code, report a problem or contribute.
What's next
More sources, and closer links between them, so your agent can follow a person or a project across chats, mail, meetings and notes.
The goal stays the same: you own your data and put it to work. Gather the right context, turn agreements into tasks and keep your projects moving, with the tools and the agent you choose.
How we fund the project
We develop custom AI workflows, assistants and bots built with LangGraph, and integrations for people and organisations. These projects support the ongoing development and maintenance of our open-source tools.
If you are exploring an integration for your own workflow, we would be happy to discuss it. You can reach us by email or Telegram.
Get in touch