AgentOps MCP Server

The AgentOps MCP server provides access to observability and tracing data for debugging complex AI agent runs. This adds crucial context about where the AI agent succeeds or fails.

Usage

MCP Client Configuration

Add the following to your MCP configuration file:

{
    "mcpServers": {
        "agentops-mcp": {
            "command": "npx",
            "args": ["agentops-mcp"],
            "env": {
              "AGENTOPS_API_KEY": ""
            }
        }
    }
}

Installation

Installing via Smithery

To install agentops-mcp for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @AgentOps-AI/agentops-mcp --client claude

Local Development

To build the MCP server locally:

# Clone and setup
git clone https://github.com/AgentOps-AI/agentops-mcp.git
cd mcp
npm install

# Build the project
npm run build

# Run the server
npm pack

Available Tools

auth

Authorize using an AgentOps project API key and return JWT token.

Parameters:

• api_key (string): Your AgentOps project API key

Returns:

• Authorization headers object or error message

get_project

Get project information and configuration.

get_trace

Retrieve trace information by ID.

Parameters:

• trace_id (string): The trace ID to retrieve

get_span

Get span information by ID.

Parameters:

• span_id (string): The span ID to retrieve

get_trace_metrics

Get performance metrics for a specific trace.

Parameters:

• trace_id (string): The trace ID

get_span_metrics

Get performance metrics for a specific span.

Parameters:

• span_id (string): The span ID

get_complete_trace

Get comprehensive trace information including all spans and their metrics.

Parameters:

• trace_id (string): The trace ID

Requirements

• Node.js >= 18.0.0
• AgentOps API key (passed as parameter to tools)