Skip to main content
The Streamkap MCP Server lets AI agents interact with your Streamkap infrastructure using natural language. It wraps the Streamkap REST API as Model Context Protocol (MCP) tools that agents can call to monitor pipeline health and metrics, inspect sources, destinations, topics, transforms, and schemas, manage the full lifecycle of your resources, and troubleshoot issues through logs and consumer group analysis.

​
Prerequisites

  • A Streamkap account
  • An API token (Client ID and Client Secret) — see API Tokens for how to create one
  • Local mode only: Node.js 20+

​
Setup Modes

The MCP server can run in two modes:
  • Remote (recommended): Connect to the hosted server at https://mcp.streamkap.com/mcp. No local installation required — credentials are passed via HTTP headers.
  • Local (via npx): Run the server locally as a child process using npx -y @streamkap/tools. Credentials are passed as environment variables. Requires Node.js 20+.

​
Setup

Remote (recommended):
claude mcp add --scope user \
  --header "X-Streamkap-Client-ID: your-client-id" \
  --header "X-Streamkap-Client-Secret: your-client-secret" \
  --transport http \
  streamkap https://mcp.streamkap.com/mcp
Local (via npx):
claude mcp add --scope user streamkap \
  -e STREAMKAP_CLIENT_ID=your-client-id \
  -e STREAMKAP_CLIENT_SECRET=your-client-secret \
  -- npx -y @streamkap/tools
For other MCP-compatible clients, use either:
  • Remote: Server URL https://mcp.streamkap.com/mcp with X-Streamkap-Client-ID and X-Streamkap-Client-Secret headers
  • Local: Command npx -y @streamkap/tools with STREAMKAP_CLIENT_ID and STREAMKAP_CLIENT_SECRET environment variables

​
Verify the Connection

After setup, verify it works by asking your AI agent:
  • “Give me an overview of my infrastructure”
  • “List all my pipelines and their health”
  • “Are any of my sources broken? Show me the details”
  • “Check the logs for any errors in the last hour”
  • “Show me the metrics for my destinations”
Once connected, your agent can look into statistics, fetch metrics, manage entities like sources, destinations, pipelines, and transforms, query logs, inspect topics, and more. For details on specific APIs, see the API Reference.

​
How Authentication Works

Your Client ID and Client Secret are passed with each session — as HTTP headers in remote mode, or as environment variables in local mode. The MCP server exchanges them for a short-lived JWT token and handles automatic refresh — no manual token management required. Sensitive fields (passwords, keys) are never exposed in tool responses. See API Tokens for a step-by-step guide on creating your Client ID and Client Secret.
Keep your Client Secret safe. If compromised, delete the token from the API Tokens page and create a new one.
  • Agents — overview of all agent integration paths
  • CLI — command-line tool with agent-friendly JSON output and scripting support
  • API Reference — full REST API documentation