Tracing Claude Code

MLflow Tracing provides automatic tracing for Claude Code:

1. CLI tracing: Automatically trace interactive Claude Code CLI conversations
2. SDK tracing: Trace Claude Agent SDK usage in Python applications

After setting up auto tracing, MLflow will automatically capture traces of your Claude Code conversations and log them to the active MLflow experiment. The trace automatically captures information such as:

• User prompts and assistant responses
• Tool usage (file operations, code execution, web searches, etc.)
• Conversation timing and duration
• Tool execution results
• Session metadata including working directory and user

Setup

Claude Code tracing can be configured using either CLI commands (for interactive use) or Python SDK imports (for programmatic use).

CLI Tracing

CLI Tracing Setup

Use CLI tracing to automatically capture your interactive Claude Code CLI conversations.

Requirements

• MLflow >= 3.4 (pip install mlflow>=3.4)
• Claude Code CLI installed and configured

Basic Setup

# Set up tracing in current directory
mlflow autolog claude

# Set up tracing in specific directory
mlflow autolog claude ~/my-project

# Check tracing status
mlflow autolog claude --status

# Disable tracing
mlflow autolog claude --disable

Configuration Examples

# Set up with custom tracking URI
mlflow autolog claude -u file://./custom-mlruns
mlflow autolog claude -u sqlite:///mlflow.db

# Set up with Databricks backend and a specific experiment ID
mlflow autolog claude -u databricks -e 123456789

# Set up with specific experiment
mlflow autolog claude -n "My AI Project"

How It Works

1. Setup Phase: The mlflow autolog claude command configures Claude Code hooks in a .claude/settings.json file in your project directory
2. Automatic Tracing: When you use the claude command in the configured directory, your conversations are automatically traced
3. View Results: Use the MLflow UI to explore your traces

Basic Example

# Set up tracing in your project
mlflow autolog claude ~/my-project

# Navigate to project directory
cd ~/my-project

# Use Claude Code normally - tracing happens automatically
claude "help me refactor this Python function to be more efficient"

# View traces in MLflow UI
mlflow ui

SDK Tracing

SDK Tracing Setup

Use SDK tracing when building applications that programmatically use the Claude Agent SDK.

Requirements

• MLflow >= 3.5 (pip install mlflow>=3.5)
• Claude Agent SDK >= 0.1.0 (pip install claude-agent-sdk >= 0.1.0)

Basic Setup

import mlflow.anthropic

# Enable automatic tracing for Claude Agent SDK
mlflow.anthropic.autolog()

Once enabled, all Claude Agent SDK interactions will be automatically traced.

note

Only ClaudeSDKClient supports tracing. Directly calling query will not be traced!

Complete Example

import asyncio
import mlflow.anthropic
from claude_agent_sdk import ClaudeSDKClient

# Enable autologging
mlflow.anthropic.autolog()

# Optionally configure MLflow experiment
mlflow.set_experiment("my_claude_app")


async def main():
    async with ClaudeSDKClient() as client:
        await client.query("What is the capital of France?")

        async for message in client.receive_response():
            print(message)


if __name__ == "__main__":
    asyncio.run(main())

Claude Tracing with MLflow GenAI Evaluation

You can also use SDK tracing with MLflow's GenAI evaluation framework:

import asyncio
import pandas as pd
from claude_agent_sdk import ClaudeSDKClient

import mlflow.anthropic
from mlflow.genai import evaluate, scorer
from mlflow.genai.judges import make_judge

mlflow.anthropic.autolog()


async def run_agent(query: str) -> str:
    """Run Claude Agent SDK and return response"""
    async with ClaudeSDKClient() as client:
        await client.query(query)

        response_text = ""
        async for message in client.receive_response():
            response_text += str(message) + "\n\n"

        return response_text


def predict_fn(query: str) -> str:
    """Synchronous wrapper for evaluation"""
    return asyncio.run(run_agent(query))


relevance = make_judge(
    name="relevance",
    instructions=(
        "Evaluate if the response in {{ outputs }} is relevant to "
        "the question in {{ inputs }}. Return either 'pass' or 'fail'."
    ),
    model="openai:/gpt-4o",
)

# Create evaluation dataset
eval_data = pd.DataFrame(
    [
        {"inputs": {"query": "What is machine learning?"}},
        {"inputs": {"query": "Explain neural networks"}},
    ]
)

# Run evaluation with automatic tracing
mlflow.set_experiment("claude_evaluation")
evaluate(data=eval_data, predict_fn=predict_fn, scorers=[relevance])

Troubleshooting

CLI Tracing

Check CLI Status

mlflow autolog claude --status

This shows:

• Whether tracing is enabled
• Current tracking URI
• Configured experiment
• Any configuration issues

Common CLI Issues

Tracing not working:

• Ensure you're in the configured directory
• Check that .claude/settings.json exists
• Review logs in .claude/mlflow/claude_tracing.log

Missing traces:

• Check if MLFLOW_CLAUDE_TRACING_ENABLED=true in your configuration
• Verify the tracking URI is accessible
• Review logs in .claude/mlflow/claude_tracing.log

Disable CLI Tracing

To stop automatic CLI tracing:

mlflow autolog claude --disable

This removes the hooks from .claude/settings.json but preserves existing traces.

SDK Tracing

Common SDK Issues

No traces appearing:

• Tracing only works with ClaudeSDKClient - direct usage of query() is not supported
• Verify mlflow.anthropic.autolog() is called before creating the ClaudeSDKClient
• Check that the tracking URI and experiment ID are configured correctly

Disable SDK Tracing

To disable SDK tracing:

mlflow.anthropic.autolog(disable=True)