home / mcp / icon mcp v105 mcp server

ICON mcp v105 MCP Server

Provides access to ICON mcp v105 endpoints via MCP for AI agents to interact with the API.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "traia-io-icon-mcp-v105-mcp-server": {
      "url": "http://localhost:8000/mcp/",
      "headers": {
        "PORT": "8000",
        "STAGE": "MAINNET",
        "LOG_LEVEL": "INFO"
      }
    }
  }
}

You connect to the ICON mcp v105 API through a dedicated MCP server that exposes the endpoints you need via a stable, protocol-wide interface. This server lets AI agents and large language models interact with ICON mcp v105 data and actions using standardized tools, enabling automated workflows, data retrieval, and operations across the ICON ecosystem.

How to use

Connect to the MCP server using your preferred MCP client by pointing it at the connection URL exposed by the server. You can call available tools to perform actions, query data, and receive structured results that you can feed into downstream AI workflows. Start by health-checking the server to confirm it is ready, then list supported tools to see what you can invoke. When you perform an action, you’ll receive a structured response that you can parse and use in your automation pipelines.

How to install

Prerequisites: you need Python and a modern runtime environment, plus access to a container runtime if you choose Docker. You will also use a lightweight MCP client to connect to the server.

# Option A: Docker (recommended for quick start) 
./run_local_docker.sh

# Option B: Local Python run (manual install)
uv pip install -e .
uv run python -m server

Configuration and environment

Configure runtime environment and keys through the supported environment variables. The server exposes a standard port and stage control to adapt to different environments.

Troubleshooting

If the server fails to start, check the container or process logs for errors. Ensure the port you expect is open and that the service is listening on the correct path. For connectivity issues, verify that you can reach the MCP URL and that the server is healthy before attempting tool calls.

Notes on usage with CrewAI

You can connect to the MCP server from your CrewAI setup and enumerate available tools. Then you can invoke a tool (for example, example_tool) to perform a predefined action and handle the response within your code.

Security considerations

Keep your environment secure by rotating credentials regularly and limiting access to the MCP endpoint. Use the standard deployment isolation mechanisms offered by your hosting environment and monitor logs for unusual activity.

Available tools

example_tool

Placeholder tool (to be implemented)