For AI agents

Give Your AI Agents the Skill to Find B2B Leads.

LeadRM turns your assistant into a data engine. Agents search your saved leads, kick off live directory scrapes on Google Maps and Yelp, and push businesses through AI decision-maker enrichment — all authenticated with a single API key you generate in the dashboard.

Get your API key View OpenAPI Schema

Claude & IDEs (Cursor / Windsurf)

The official LeadRM MCP server plugs into Claude Desktop, Cursor, Windsurf, Copilot, and Gemini over stdio. Your agent getsquery_leads,trigger_directory_scrape, andenrich_leads_batchtools plus a live workspace-stats resource.

npx -y leadrm-mcp

Set LEADRM_API_KEY in your client's env. One key per user — revoke anytime from Settings.

ChatGPT & OpenAI Actions

Point a Custom GPT or OpenAI agent at our OpenAPI 3.1 schema. TheGET /api/v1/agent/leadsendpoint accepts a Bearer API key and filters byindustry,city, andhas_email— returning leads as clean JSON, including enriched decision-makers.

View OpenAPI Schema

Custom Agents (CrewAI / LangChain)

Build programmatic pipelines with outbound webhooks and API keys. Push enriched leads straight into CrewAI crews, LangChain tools, Pabbly Connect, Zapier, Make, GoHighLevel, or HubSpot — your agent stack stays the orchestrator.

GET /api/v1/agent/leads?industry=hvac&city=austin

Configure webhooks & keys

How authentication works

Open Settings → Integrations → MCP server access and generate a key (lrmk_…). It is shown once — only a SHA-256 hash is stored.
Send it as Authorization: Bearer to the REST endpoint, or set it as LEADRM_API_KEY for the MCP server.
Agents only ever see your own workspace leads — the same security rules as the app. Credits, BYOK, and rate limits apply exactly as they do in the dashboard.