Glossary
What is the Model Context Protocol (MCP)?
Updated 1 min read By Operators team
Anthropic introduced MCP in November 2024, and other assistant makers have since adopted it. Before it, every pairing of an assistant and a tool needed its own integration. With it, a tool publishes one server address and assistants that speak the protocol can use it.
What matters in practice is the permission step: which data the assistant may see, and whether it may change anything. Operators offers an MCP server address for your board. Claude or your own scripts can read and, once you untick Read only, add items in the companies you choose. Whether ChatGPT can make changes depends on your ChatGPT plan; on plans that cannot, it reads.
Related
- Guides How to connect Claude or ChatGPT to your Operators board Connect Claude, Claude Code or ChatGPT to your Operators board with one link, choose which companies it may see, and decide if it may make changes. 4 min read
- Guides How to let an AI agent work on your board safely Give AI agents room to work on your Operators board with per-company access, read-only keys, approval requests and one-click undo for every change. 5 min read
- Articles Read-only by default, a sane rule for connecting tools Why every new connection between your tools, apps and AI assistants should start read only, when to allow changes, and what to check before approving. 5 min read