Tracxn MCP follows the open MCP specification, so it works with any compliant client. Common ones include Claude Desktop, Cline (VS Code), Continue, Windsurf, and custom clients built with the MCP SDK.
Generic configuration
Most MCP clients accept this configuration format:
json
{
"mcpServers": {
"tracxn": {
"url": "https://platform.tracxn.com/mcp"
}
}
}
The client will detect the OAuth requirement and prompt you to log in via browser on first use.
Building a custom integration
If you're building your own AI agent or internal LLM tool, you can integrate Tracxn MCP using any MCP client library:
Python: mcp Python SDK
TypeScript: @modelcontextprotocol/sdk
Other languages: See the MCP specification
Point your client at the Tracxn MCP URL above. OAuth flow is handled by the SDK — your end user authenticates once and tokens are managed for you.
For enterprise customers integrating Tracxn MCP into internal LLM tools, contact your account manager for guidance on auth flows, scoping, and best practices.