Skip to content

Mcp discovery

MCPDiscoveryTool #

Bases: BaseToolSpec

MCP Discovery Tool.

This tool queries the MCP Discovery API for autonomous tool recommendations. It accepts a natural language description of the need and returns a human-readable list of recommended MCP servers with name, category, and description.

Attributes:

Name Type Description
api_url

The URL of the MCP discovery API endpoint.
Source code in .build/python/llama-index-integrations/tools/llama-index-tools-mcp-discovery/llama_index/tools/mcp_discovery/base.py 
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
class MCPDiscoveryTool(BaseToolSpec):
    """
    MCP Discovery Tool.

    This tool queries the MCP Discovery API for autonomous tool recommendations.
    It accepts a natural language description of the need and returns a
    human-readable list of recommended MCP servers with name, category, and description.

    Attributes:
        api_url: The URL of the MCP discovery API endpoint.

    """

    spec_functions = ["discover_tools"]

    def __init__(self, api_url: str) -> None:
        """
        Initialize the MCP Discovery Tool.

        Args:
            api_url: The URL of the MCP discovery API endpoint.

        """
        self.api_url = api_url

    async def discover_tools(self, user_request: str, limit: int = 5) -> str:
        """
        Discover tools based on a natural language request.

        This method allows an agent to discover needed tools without human intervention.
        It queries the MCP discovery API with the user's request and returns formatted
        tool recommendations.

        Args:
            user_request: Natural language description of the tool needed.
            limit: Maximum number of tool recommendations to return. Defaults to 5.

        Returns:
            A formatted string containing the discovered tools with their names,
            descriptions, and categories. Returns an error message if the request fails.

        Example:
            >>> tool = MCPDiscoveryTool(api_url="http://localhost:8000/api")
            >>> result = await tool.discover_tools("I need a math calculator", limit=3)
            >>> print(result)
            Found 2 tools:
            1. Name: math-calculator,
               Description: A tool for calculations,
               Category: math

        """
        try:
            async with aiohttp.ClientSession() as session:
                async with session.post(
                    self.api_url, json={"need": user_request, "limit": limit}
                ) as response:
                    data = await response.json()

            tools_json = data.get("recommendations", [])
            num = data.get("total_found", -1)

            if num == -1:
                tools = "Following tools are found:\n"
            else:
                tools = f"Found {num} tools:\n"

            if tools_json:
                for ind, i in enumerate(tools_json, start=1):
                    tools += f"{ind}. Name: {i.get('name')},\n"
                    tools += f"   Description: {i.get('description')},\n"
                    tools += f"   Category: {i.get('category')}\n\n"
                return tools.strip()

            return tools

        except Exception as e:
            return f"Error discovering tools: {e}"

discover_tools async #

discover_tools(user_request: str, limit: int = 5) -> str

Discover tools based on a natural language request.

This method allows an agent to discover needed tools without human intervention. It queries the MCP discovery API with the user's request and returns formatted tool recommendations.

Parameters:

Name Type Description Default
user_request str

Natural language description of the tool needed.

 required
limit int

Maximum number of tool recommendations to return. Defaults to 5.

 5

Returns:

Type Description
str

A formatted string containing the discovered tools with their names,
str

descriptions, and categories. Returns an error message if the request fails.
Example

tool = MCPDiscoveryTool(api_url="http://localhost:8000/api") result = await tool.discover_tools("I need a math calculator", limit=3) print(result) Found 2 tools: 1. Name: math-calculator, Description: A tool for calculations, Category: math

Source code in .build/python/llama-index-integrations/tools/llama-index-tools-mcp-discovery/llama_index/tools/mcp_discovery/base.py 
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
async def discover_tools(self, user_request: str, limit: int = 5) -> str:
    """
    Discover tools based on a natural language request.

    This method allows an agent to discover needed tools without human intervention.
    It queries the MCP discovery API with the user's request and returns formatted
    tool recommendations.

    Args:
        user_request: Natural language description of the tool needed.
        limit: Maximum number of tool recommendations to return. Defaults to 5.

    Returns:
        A formatted string containing the discovered tools with their names,
        descriptions, and categories. Returns an error message if the request fails.

    Example:
        >>> tool = MCPDiscoveryTool(api_url="http://localhost:8000/api")
        >>> result = await tool.discover_tools("I need a math calculator", limit=3)
        >>> print(result)
        Found 2 tools:
        1. Name: math-calculator,
           Description: A tool for calculations,
           Category: math

    """
    try:
        async with aiohttp.ClientSession() as session:
            async with session.post(
                self.api_url, json={"need": user_request, "limit": limit}
            ) as response:
                data = await response.json()

        tools_json = data.get("recommendations", [])
        num = data.get("total_found", -1)

        if num == -1:
            tools = "Following tools are found:\n"
        else:
            tools = f"Found {num} tools:\n"

        if tools_json:
            for ind, i in enumerate(tools_json, start=1):
                tools += f"{ind}. Name: {i.get('name')},\n"
                tools += f"   Description: {i.get('description')},\n"
                tools += f"   Category: {i.get('category')}\n\n"
            return tools.strip()

        return tools

    except Exception as e:
        return f"Error discovering tools: {e}"

options: members: - MCPDiscoveryTool