Server: Tools MCP

Deploy Python tools as MCP servers for Claude Desktop, Cursor, and other MCP clients

Expose Python functions as MCP (Model Context Protocol) tools for Claude Desktop, Cursor, and other MCP clients.

CLI

pip install "praisonaiagents[mcp]"

# Start MCP server
praisonai serve mcp --transport sse --port 8080

# Start with specific server name
praisonai serve mcp --name my-server --transport sse --port 8080

Option	Default	Description
--host	127.0.0.1	Host to bind to
--port	8080	Server port
--transport	stdio	Transport: stdio, sse
--name	-	Server name from config

Python - Custom Tools

from praisonaiagents import ToolsMCPServer

def search(query: str) -> str:
    """Search the web for information."""
    return f"Results for: {query}"

def calculate(expression: str) -> str:
    """Calculate a math expression."""
    return str(eval(expression))

server = ToolsMCPServer(name="my-tools")
server.register_tool(search)
server.register_tool(calculate)
server.run(transport="sse", host="0.0.0.0", port=8080)

Expected Output:

🚀 Starting MCP server 'my-tools' with SSE transport
📡 SSE endpoint: http://0.0.0.0:8080/sse
🛠️  Available tools: search, calculate

Verify:

curl http://localhost:8080/sse

Python - Built-in Tools

from praisonaiagents import launch_tools_mcp_server

launch_tools_mcp_server(
    tool_names=["tavily_search", "wikipedia_search"],
    transport="sse",
    port=8080
)

Expected Output:

🚀 Starting MCP server 'praisonai-tools' with SSE transport
📡 SSE endpoint: http://0.0.0.0:8080/sse
🛠️  Available tools: tavily_search, wikipedia_search

Python - Stdio Transport

For Claude Desktop local integration:

from praisonaiagents import ToolsMCPServer

def my_tool(query: str) -> str:
    """Process a query."""
    return f"Processed: {query}"

server = ToolsMCPServer(name="my-tools", tools=[my_tool])
server.run(transport="stdio")

Expected Output:

🚀 Starting MCP server 'my-tools' with stdio transport
🛠️  Available tools: my_tool

agents.yaml

Tools MCP server is Python-only. No agents.yaml support.

MCP Endpoints (SSE Transport)

Endpoint	Method	Description
/sse	GET	SSE connection for MCP
/messages/	POST	Send MCP messages

ToolsMCPServer Parameters

Parameter	Type	Default	Description
name	str	praisonai-tools	Server name
tools	list	None	Initial tools to register
debug	bool	False	Enable debug logging

launch_tools_mcp_server Parameters

Parameter	Type	Default	Description
tools	list	None	Custom tool functions
tool_names	list	None	Built-in tool names
name	str	praisonai-tools	Server name
transport	str	stdio	stdio or sse
host	str	0.0.0.0	Host for SSE
port	int	8080	Port for SSE

Connect MCP Client

Claude Desktop (claude_desktop_config.json):

For SSE transport:

{
  "mcpServers": {
    "my-tools": {
      "url": "http://localhost:8080/sse"
    }
  }
}

For stdio transport:

{
  "mcpServers": {
    "my-tools": {
      "command": "python",
      "args": ["/path/to/mcp_server.py"]
    }
  }
}

Cursor (.cursor/mcp.json):

{
  "mcpServers": {
    "my-tools": {
      "url": "http://localhost:8080/sse"
    }
  }
}

Test MCP Tools

# List tools
curl -X POST http://localhost:8080/messages/ \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc": "2.0", "method": "tools/list", "id": 1}'

Response:

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "tools": [
      {
        "name": "search",
        "description": "Search the web for information.",
        "inputSchema": {
          "type": "object",
          "properties": {"query": {"type": "string"}},
          "required": ["query"]
        }
      }
    ]
  }
}

# Call tool
curl -X POST http://localhost:8080/messages/ \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {"name": "search", "arguments": {"query": "AI news"}},
    "id": 2
  }'

Response:

{
  "jsonrpc": "2.0",
  "id": 2,
  "result": {
    "content": [{"type": "text", "text": "Results for: AI news"}]
  }
}

Troubleshooting

Issue	Fix
Port in use	lsof -i :8080
Missing deps	pip install "praisonaiagents[mcp]"
Tool not found	Check tool is registered before run()
SSE not connecting	Check firewall, use host="0.0.0.0"

Related

• MCP API Reference - Full MCP endpoint documentation
• Agents Server - Deploy agents as HTTP server
• A2A Server - Deploy as A2A server