> ## Documentation Index
> Fetch the complete documentation index at: https://praison.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# MCP Server • AI Agents Framework

> PraisonAI MCP Server Module

# mcp\_server

<Badge color="purple">AI Agents Framework</Badge>

PraisonAI MCP Server Module

Exposes PraisonAI capabilities as an MCP server that any MCP client can connect to.

MCP Protocol Version: 2025-11-25

Supports:

* STDIO transport (default, for Claude Desktop, Cursor, etc.)
* HTTP Stream transport (MCP 2025-11-25 spec)
* Recipe-to-MCP server bridge
* Tasks API (experimental)
* Elicitation (form and URL modes)
* OAuth 2.1 / OpenID Connect authentication
* Icons and rich metadata

Usage:

# STDIO mode (for Claude Desktop config)

praisonai mcp serve --transport stdio

# HTTP Stream mode

praisonai mcp serve --transport http-stream --port 8080

# Serve a recipe as MCP server

praisonai mcp serve-recipe support-reply --transport stdio

# Programmatic usage

from praisonai.mcp\_server import MCPServer, RecipeMCPAdapter

server = MCPServer()
server.run(transport="stdio")

# Recipe as MCP server

adapter = RecipeMCPAdapter("support-reply")
adapter.load()
server = adapter.to\_mcp\_server()
server.run(transport="stdio")

## Import

```python theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
from praisonai import mcp_server
```

***

## Related Documentation

<CardGroup cols={2}>
  <Card title="MCP Concept" icon="server" href="/docs/concepts/mcp" />

  <Card title="MCP Lifecycle" icon="rotate" href="/docs/features/mcp-lifecycle" />
</CardGroup>