Judgment Labs Logo
DocumentationSDK ReferenceChangelog
PythonIntegrations

Claude Agent SDK

setup_claude_agent_sdk()

Patch the Claude Agent SDK for automatic tracing.

Monkey-patches ClaudeSDKClient and the standalone query() function so every agent turn, LLM call, and tool invocation is recorded as a span. Call this once at startup, before any SDK usage.

from judgeval import Tracer
from judgeval.integrations import setup_claude_agent_sdk

Tracer.init(project_name="my-agent")
setup_claude_agent_sdk()

# All Claude Agent SDK calls are now traced automatically
def setup_claude_agent_sdk() -> bool:

Returns

bool - True if patching succeeded, False on error.

Langgraph

Integration that routes LangGraph spans through the Judgment pipeline.

Openlit

Integration that routes OpenLit instrumentation through Judgment.

Last updated on

On this page

setup_claude_agent_sdk()Returns