Reference | AI & MCP | Kick Docs

Kick MCP Reference Documentation

For the complete documentation index, see llms.txt. This page is also available as Markdown.

Use this page when you need exact MCP connection values or want to confirm how Kick MCP chooses credentials, workspaces, and available tools. For step-by-step setup, see Connect custom AI tools.

Hosted MCP endpoint

Remote MCP clients connect to:

https://use.kick.co/mcp

OAuth-capable clients can discover Kick's protected resource metadata at:

https://use.kick.co/.well-known/oauth-protected-resource

Kick MCP has two kinds of tools:

If your AI client uses OAuth, it requests these as the mcp:read and mcp:write scopes.

PAT bearer header

For PAT-based clients, send the token as a bearer header:

Authorization: Bearer kick_pat_...

Keep PATs out of prompts, source control, and shared docs. Revoke any token you suspect has been exposed.

Workspace selection

Workspace-scoped MCP tools can receive workspaceId in tool arguments.

Workspace-bound PATs auto-inject their bound workspace. The backend injects a default workspaceId only when the credential resolves to exactly one workspace. User-scoped PATs and OAuth tokens should pass workspaceId explicitly.

Find available tools

Use the tool list shown by your AI client as the current source of truth. Available Kick tools can vary by token scope, workspace permissions, enabled features, and client cache state.

Common ways to inspect tools:

List the Kick tools you can use. Identify which ones are read-only, then confirm the active workspace before looking up any client data.

If a tool was recently added or renamed, restart or reconnect the MCP client so it reloads the tool list.

Common tool examples

These examples are not a complete list. Use your MCP client's tool list for the up-to-date names available to your credential.

Kick MCP tools use portable names, usually lower snake case. The available list can vary by token scope, workspace permissions, and feature flags.

Safety reminders

For setup examples, see Connect custom AI tools. For more safety detail, see Permissions and Security.