MCP server + REST API

GitHub intent signals for AI agents and Clay workflows.

Q: What does the LeadCognition MCP server do?

The LeadCognition MCP server exposes GitHub-based developer intent signals to MCP clients and REST workflows. It can list Discover sessions, rank developers and companies showing intent, return company intelligence, and prepare Clay-ready exports.

Q: Can I use LeadCognition with Clay?

Yes. LeadCognition provides REST endpoints for Clay source tables, Clay enrichment, export preparation, and webhook delivery. The same API key can also be used with MCP clients.

Q: Which MCP clients are supported?

LeadCognition supports MCP-compatible clients through the @leadcognition/mcp stdio wrapper and the hosted streamable HTTP endpoint. Common use cases include Claude Desktop, Cursor, Windsurf, ChatGPT, and custom agent workflows.

LeadCognition turns public GitHub activity into ranked developers, company intent, Clay-ready rows, and API responses your GTM stack can use immediately.

Start with 50 free credits Read setup guide

Claude Desktop
Cursor
Clay source tables
REST + OpenAPI

What it gives your GTM team

One developer-intent layer, three integration paths.

Use the hosted MCP endpoint for agent workflows, REST endpoints for enrichment pipelines, or Clay-specific export routes when your team wants spreadsheet-shaped rows.

MCP for agents

Connect Claude Desktop, Cursor, Windsurf, ChatGPT, or a custom agent to query company intelligence and developer intent using bearer auth.

Clay-ready exports

Source domain-driven rows through /api/v1/clay/source, enrich a company, or send rows to a webhook with mcp:export.

REST for RevOps

Use the OpenAPI 3.1 contract for scripts, n8n, internal tools, or partner integrations.

Self-serve trial

Get an API key and test real developer signals.

The trial creates a workspace-linked key with mcp:read and mcp:export scopes. After activation, paste it into Clay, Claude Desktop, Cursor, or your REST client.

1

Request a magic link

Use a work email so the trial can attach to the right company domain.
2

Activate with GitHub or Google

The activation step keeps API keys tied to a real LeadCognition workspace.
3

Connect your workflow

Follow this MCP setup guide, import the OpenAPI spec, or start from a Clay source table.

Quick start after activation

curl -X POST https://app.leadcognition.io/api/v1/clay/source \
  -H "Authorization: Bearer lc_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain": "vercel.com", "limit": 10}'

Available tools

Signals your agent can ask for.

Company intelligence

get_company_intel returns developer count, signal volume, top repositories, and the next best GTM action for a company domain.

Developer intent

discover_developer_intent and list_developer_signals rank developers by recent GitHub activity, repo fit, and engagement strength.

Pipeline export

prepare_clay_export and send_to_clay turn signal rows into the format your enrichment or outreach workflow expects.

MCP FAQ

Answers for AI search and human buyers.

What does the LeadCognition MCP server do?

It exposes GitHub-based developer intent signals to MCP clients and REST workflows, including company intelligence, developer rankings, Discover session results, and Clay exports.

Can I use LeadCognition with Clay?

Yes. LeadCognition has REST endpoints for Clay source tables, Clay enrichment, export preparation, and webhook delivery.

Which MCP clients are supported?

Use the @leadcognition/mcp stdio wrapper with Claude Desktop, Cursor, Windsurf, and other clients, or call the hosted streamable HTTP endpoint directly.