home / mcp / toolresults mcp server

Toolresults MCP Server

MCP Server generated by mcp.ag2.ai

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "ag2-mcp-servers-cloud-tool-results-api": {
      "command": "python",
      "args": [
        "mcp_server/main.py",
        "stdio"
      ],
      "env": {
        "CONFIG": "{...config...}",
        "SECURITY": "YOUR_API_KEY",
        "CONFIG_PATH": "mcp_server/mcp_config.json"
      }
    }
  }
}

You run an MCP server that exposes a programmable interface for managing and querying model context data. This server helps you connect your MCP clients to the Tool Results API workflow, enabling standardized communication, testing, and automation across tools and services.

How to use

Start the MCP server in stdio mode to run locally and connect with an MCP client. The server exposes a standard transport that your client can interact with using the MCP protocol. You can configure transport, security, and the location of your configuration at startup to fit your environment.

To interact with the server, initiate the standard stdio transport from your client workflow. You typically launch the server process and then direct your MCP client to send requests through the established transport channel. Use the following starting point as you integrate your client: the server is started via a Python script and accepts stdio as the transport.

How to install

Prerequisites: ensure you have Python 3.9+ installed on your system. You also need pip and uv to run and manage the server in development or lightweight environments.

Step 1: Prepare the server runtime
- Ensure you have Python 3.9 or newer installed
- Confirm that pip and uv are available on your system

Step 2: Run the MCP server locally in stdio mode - Start the server with Python by invoking the main script in stdio mode

python mcp_server/main.py stdio

Additional configuration and startup notes

Configuration can be supplied at startup through environment variables. The following variables are recognized for startup and security purposes:

CONFIG_PATH: Path to a JSON configuration file (for example, mcp_server/mcp_config.json)
CONFIG: A JSON string containing the configuration
SECURITY: Environment variables for security parameters (for example, API keys)

Security and configuration details

To tailor the server to your environment, provide a configuration JSON via CONFIG_PATH or CONFIG. If you use CONFIG_PATH, point it to the file that contains your MCP configuration. If you use CONFIG, pass the JSON inline. For secured deployments, populate SECURITY with your API keys or other required credentials per your organization's policy.

Available tools

lint

Run linting with ruff to enforce code quality and consistency.

static-analysis

Perform static analysis with mypy, bandit, and semgrep to catch type issues, security risks, and code patterns.

tests

Execute tests with pytest to verify functionality and coverage.

server-run

Start and interact with the MCP server using stdio transport via the main Python entry point.