Getting Started

Retrace captures every LLM call, tool invocation, and error your AI agent makes, then lets you replay, fork, and share the execution like a video.

Quick Start

1. Install the SDK

pip install retrace

2. Configure your API key

import retrace

retrace.configure(api_key="rt_live_...")

3. Add the decorator

@retrace.record(name="my-agent")
def run_agent(prompt: str):
    response = client.chat.completions.create(
        model="gemini-3.1-pro-preview",
        messages=[{"role": "user", "content": prompt}]
    )
    return response.choices[0].message.content

4. Verify in the dashboard

Run your agent. Open the Retrace dashboard to see the recorded trace with every step visualized.

What Gets Captured

[!TIP] Set RETRACE_ENABLED=false to disable recording in production without changing code.

Next Steps

• Recording Traces - Deep dive into the @record decorator
• Replaying Executions - Learn the tape player
• Forking - Branch and replay with modifications
• API Reference - Full REST API documentation