SIGNALRAVEN
DEVELOPERS · API & MCP

Build on SignalRaven.

Pull your buying-intent signals into your own stack with the REST API, or connect them to Claude and ChatGPT over MCP. Two surfaces, one OAuth, scoped to your workspace.

Two surfaces, one auth

  • REST API — api.signalraven.ai/api/v1 — for your own backend, scripts, and data pipelines.
  • MCP server — api.signalraven.ai/mcp — for AI assistants like Claude and ChatGPT.

Both authenticate with OAuth 2.1 via auth.signalraven.ai and return the same workspace-scoped data. Reads are unrestricted; anything that spends credits or changes where signals are delivered sits behind its own write scope, named on the consent screen. Nothing on either surface can touch a LinkedIn account.

REST API
API reference →
Endpoints, scopes, auth, examples.
MCP · REFERENCE
MCP server →
Tools, transport, discovery — for building against it.
MCP · SET UP
Connect Claude / ChatGPT →
Add the connector and start asking — no code.
How to connect to the SignalRaven API and MCP: your clients authenticate with OAuth at auth.signalraven.ai, then call the REST API or MCP server for your workspace data.
Your clients → authenticate → SignalRaven surface → your data

Get started

  1. Create an API key — app.signalraven.ai→ Settings → API keys (workspace admin). You’ll get a client id + secret once.
  2. Exchange it for a token at auth.signalraven.ai/oauth2/token using the OAuth client_credentials grant.
  3. Call the API or MCP server with the bearer token. See the REST reference or MCP reference.
All developer docs