Developer Hub

Octolane is programmable

MCP server for AI agents, REST API for custom integrations, webhooks for real-time events. Build on top of your CRM instead of around it.

1Beta

MCP Server

Connect Octolane to Claude Desktop, Cursor, Claude Code, or any MCP-compatible AI agent. Query your pipeline, create deals, draft follow-ups, all through natural language. No other CRM can do this.

Show me all deals closing this quarterNatural language pipeline queries
Create a deal for Vercel, $120k, closing JuneCreate records conversationally
Draft a follow-up for my last call with ResendTrigger actions from any agent
What's my pipeline health this month?Instant reporting, no dashboards
$ npx @octolane/mcp-server --api-key sk_live_...
2Stable

REST API

Full CRUD for contacts, deals, accounts, activities, and notes. Everything you need to build custom integrations, sync data, or automate your sales stack.

GET/contactsList, filter, and search contacts
POST/dealsCreate and update deals programmatically
GET/activitiesFetch call logs, emails, and meetings
PATCH/deals/:idUpdate deal stage, amount, or owner
GET/accountsCompany records with enrichment data
POST/notesAttach notes to any record
Full API reference
3Stable

Webhooks

Event-driven triggers for every meaningful CRM action. Wire Octolane into Slack, Zapier, n8n, or your own internal tools. Build custom workflows without polling.

deal.createdFires when a new deal is created, manually or by AI
deal.stage_changedPipeline stage transitions, with before/after values
followup.sentAuto-generated follow-up email was sent to a prospect
contact.createdNew contact added from a meeting, email, or enrichment
meeting.recordedMeeting transcript and summary are ready
deal.staleDeal marked stale after no activity for N days
{
  "event": "deal.stage_changed",
  "data": {
    "deal_id": "d_8xk2m",
    "name": "Vercel — Enterprise",
    "from_stage": "Demo",
    "to_stage": "Negotiation",
    "amount": 120000
  }
}

Most CRMs give you an API as an afterthought. Octolane is built API-first, we connected it to our AI agent in 10 minutes and it just works.