MCP server

Blitflow speaks the Model Context Protocol, so AI agents — Claude Code, Claude Desktop, Cursor, or anything MCP-capable — can discover nodes, inspect workflows, and run them as tools. The tools map one-to-one onto the API contract: same operations, same schemas.

Hosted server (recommended)

The hosted server lives at:

https://mcp.blitflow.com

It uses streamable HTTP transport and OAuth — your MCP client opens a browser to sign you in; there are no tokens to paste.

claude mcp add --transport http blitflow https://mcp.blitflow.com

For clients configured with a .mcp.json-style file:

{
  "mcpServers": {
    "blitflow": {
      "type": "http",
      "url": "https://mcp.blitflow.com"
    }
  }
}

Local server (stdio)

The SDK also ships a stdio MCP server, blitflow-mcp, which authenticates with a personal access token instead of OAuth — useful for headless environments or when pointing at a non-production API host.

Env var	Required	Description
`BLITFLOW_TOKEN`	yes	Your `blit_…` token
`BLITFLOW_API_URL`	no	Override the API base URL

{
  "mcpServers": {
    "blitflow": {
      "command": "blitflow-mcp",
      "env": {
        "BLITFLOW_TOKEN": "blit_..."
      }
    }
  }
}

What the agent can do

Tool	Scope	Purpose
`nodes_list`	`read`	Search the node palette
`nodes_get`	`read`	Read one node's exact inputs and enum values
`workflows_get`	`read`	Fetch a published workflow definition
`runs_create`	`run`	Run a workflow (inline or by `<id>@<version>`)
`runs_node`	`run`	Run a single node

See Tools for full parameter schemas, and note that run-scoped tools spend money — agents should pass maxCostUsd on every run.

Tool results come back as JSON text. On failure a tool returns an error result with the message — agents should read it and adjust, not retry blindly.

Hosted server (recommended)

Local server (stdio)

What the agent can do

On this page