Skip to main content
An agent that writes code needs somewhere real to run it. slipway exposes the same project lifecycle your team uses over a first-party Model Context Protocol server, so an agent can deploy a change, hit the URL, read the logs and verify the fix before calling the task done.

​
What an agent can do

An agent drives the same projects your team uses, through the same build, deploy and teardown path.

​
Connecting

Create a token under Settings → MCP in the console, then point your agent at the endpoint with an Authorization: Bearer header.
Connection steps for Cursor and Claude Desktop, the full tool list and the role table are on the MCP server page.

​
Scope the token

Every token carries a role, from viewer through admin, and each tool requires a minimum role. Give an agent the lowest role that lets it do its job. An agent that only inspects builds needs viewer; one that writes secrets needs admin. Tokens are scoped to a single organization, stored hashed, and revocable at any time. See roles.

​
Useful prompts

Once connected, an agent can be asked directly:
  • Deploy this branch to slipway and give me the URL.
  • The deploy failed. Read the logs and tell me why.
  • Run the test suite inside the running API service.
  • Tear down the project when you are done.