# Overview (/en/docs/api-mcp/get-started)





This is the shortest path to using the CentralCSP API: create a key, make one
authenticated request, then connect an AI agent over the [Model Context Protocol
(MCP)](https://modelcontextprotocol.io/).

## 1. Create an API key [#1-create-an-api-key]

Create an API key from your workspace settings, give it a name, and store the
returned value as a secret; treat it like a password. Each key is prefixed `ccsp_`
and is scoped to that one workspace.

<img alt="The Create API key dialog with name, expiry and allowed IPs" src="__img0" width="1359" height="645" />

## 2. Make your first request [#2-make-your-first-request]

Send the key as a bearer token with each request and call an endpoint from the
reference. Requests go to the `/v1` base path:

```bash
curl -H "Authorization: Bearer ccsp_..." \
  https://api.centralcsp.com/v1/workspaces
```

The full endpoint list is generated from the [OpenAPI](https://www.openapis.org/)
spec, see **Reference** in the sidebar.

## 3. Connect over MCP [#3-connect-over-mcp]

To let an AI assistant query your CentralCSP data, connect its MCP client to the
CentralCSP [MCP server](/en/docs/api-mcp/mcp). There is no API key in that flow,
sign-in is OAuth in the browser. The [MCP quickstart](/en/docs/api-mcp/mcp/quickstart)
is the short path from adding the server to reading your first reports.

## Next steps [#next-steps]

* Browse every endpoint in the API **Reference** (sidebar).
* Read the conceptual material (auth, rate limits, webhooks) in the [API & MCP overview](/en/docs/api-mcp).