MCP (Model Context Protocol) is how Claude Code and other agents connect to external tools and data. Add an MCP server and the agent can act on GitHub, Slack, Supabase, Notion, and thousands of other systems, not just your code.
Key takeaways
- MCP is an open standard that exposes tool actions to agents.
- Add a server and GitHub, Slack, Supabase and more become agent tools.
- Hosted agents bundle the popular connectors so you skip the wiring.
What MCP does
It's an open standard that exposes a tool's actions to an agent in a consistent way. Point Claude Code at an MCP server and those actions become available to it.
Adding a server
Configure the MCP server in Claude Code's settings; it then appears as a set of tools the agent can call. Popular ones: GitHub, Slack, Supabase, Linear, Gmail.
Skip the wiring
Creative Tim's hosted agents (OpenClaw/Hermes) come with these connectors preconfigured, so you don't have to set up MCP servers yourself, just spawn and connect.
Skip the setup, spawn a hosted agent
Creative Tim hosts Claude-Code-compatible harnesses (OpenClaw, Hermes), preconfigured with connectors. Bring your own key or run on credits.
Frequently asked questions
What is an MCP server?
A small service that exposes a tool's actions (e.g. GitHub) to an AI agent via the Model Context Protocol.
Do I have to build one?
No, many exist already, and hosted agents bundle the popular ones. You can build custom servers for proprietary systems.
Is MCP specific to Claude Code?
No. MCP is an open standard, so the same servers work across many agents and tools, including OpenClaw and Hermes.
Keep reading