Skip to main content
Tools in MCP are server-implemented functions that enable AI models to perform specific actions. They represent capabilities that a server exposes to clients, allowing AI applications to interact with external systems, execute commands, or manipulate data.

​
What are Tools?

Tools are well-defined functions with:
  • JSON Schema definitions that specify their inputs and outputs
  • Single, focused operations that perform one specific task
  • User approval requirement before execution for security
Common examples include:
  • File operations (read, write, delete)
  • Web browsing and data fetching
  • Database queries
  • API calls to external services
  • System commands

​
Listing Available Tools

To see what tools are available from a connected MCP server: 
import asyncio
from mcp_use import MCPClient

async def list_tools():
    # Initialize client with server configuration
    config = {
        "mcpServers": {
            # Your server definitions here
        }
    }
    client = MCPClient(config)

    # Connect to servers
    await client.create_all_sessions()

    # Get a session for a specific server
    session = client.get_session("my_server")

    # List all available tools - always returns fresh data
    tools = await session.list_tools()

    for tool in tools:
        print(f"Tool: {tool.name}")
        print(f"Description: {tool.description}")
        print(f"Schema: {tool.inputSchema}")
        print("---")

# Run the example
asyncio.run(list_tools())

​
Automatic Tool List Update

When servers send ToolListChangedNotification, it signals that the tool list has changed. The list_tools() method always fetches fresh data from the server, ensuring you get up-to-date information. Important: Always use await session.list_tools() instead of the deprecated session.tools property to ensure you get fresh data: 
# ✅ Recommended - always returns fresh data
tools = await session.list_tools()

# ⚠️ Deprecated - may return stale data
# tools = session.tools

​
Calling Tools

Tools are executed using the call_tool method: 
import asyncio
from mcp_use import MCPClient

async def call_tool_example():
    config = {
        "mcpServers": {
            # Your server definitions here
        }
    }
    client = MCPClient(config)
    await client.create_all_sessions()

    session = client.get_session("filesystem_server")

    # Call a tool with arguments
    result = await session.call_tool(
        name="read_file",
        arguments={
            "path": "/path/to/file.txt",
            "encoding": "utf-8"
        }
    )

    # Handle the result
    if result.isError:
        print(f"Error: {result.content}")
    else:
        print(f"File content: {result.content}")

asyncio.run(call_tool_example())

​
Tool Results

Tool calls return a CallToolResult object with:
  • content: The result data or error message
  • isError: Boolean indicating success or failure
  • Additional metadata about the execution
# Example of handling tool results
result = await session.call_tool("search_web", {"query": "MCP protocol"})

if result.isError:
    print(f"Tool execution failed: {result.content}")
else:
    # Process successful result
    for item in result.content:
        print(f"Found: {item}")

​
Error Handling

Always handle potential errors when calling tools: 
try:
    result = await session.call_tool("risky_operation", {"param": "value"})
    if result.isError:
        print(f"Tool reported error: {result.content}")
    else:
        print(f"Success: {result.content}")
except Exception as e:
    print(f"Connection or protocol error: {e}")