MCP Server

The VelaFlows MCP Server gives AI agents — like Claude, ChatGPT, Cursor, and Windsurf — direct access to your VelaFlows platform. It enables AI to search your API documentation, execute API calls, and build workflow automations on your behalf.

What is MCP?

The Model Context Protocol (MCP) is an open standard that lets AI assistants connect to external tools and data sources. Instead of copying and pasting API docs into your chat, the AI can query VelaFlows directly — searching endpoints, reading node configurations, and even making live API calls.

What’s Included

The VelaFlows MCP server provides 36 capabilities:

Category	Count	Description
Tools	22	Execute actions — search APIs, make requests, build workflows
Resources	8	Read-only context — API catalog, node reference, event catalog, guides
Prompts	6	Pre-built templates — integration plans, code generation, debugging

Two Ways to Connect

Mode	Best For	Setup
Local (CLI)	Claude Code, Cursor, Windsurf, VS Code	`npm install -g @velaflows/mcp-server`
Hosted (SSE)	Claude Desktop, ChatGPT Desktop	Connect to `https://mcp-server-production-c72e.up.railway.app/sse`

How It Works


You: "List all VelaFlows services with their endpoint counts"

AI Agent:
  → calls list-services tool
  → returns 34 services, 1,093 total endpoints

You: "Search for endpoints related to sending messages"

AI Agent:
  → calls search-api tool with query "send message"
  → returns matching endpoints from Inbox, Channels, Broadcast services

You: "Create a workflow that sends a Slack notification when a new lead is created"

AI Agent:
  → calls list-events to find "crm.lead.created" trigger
  → calls get-node-details for "notify-slack" node config
  → calls suggest-workflow to generate the DAG
  → calls create-workflow to save it

Authentication

Reference tools (search, list, get details) work without an API token — they query bundled documentation data
Live API tools (api-get, api-post, create-workflow, etc.) require a VelaFlows API token (sk_live_...)
Create a token in your VelaFlows dashboard under Developer → API Tokens

Quick Start

Fastest: Global Install


npm install -g @velaflows/mcp-server
velaflows-mcp-server

The server starts in stdio mode. Press Ctrl+C to stop.