Instrument

Add observability to your AI agents with automatic and manual instrumentation.

The Instrument section covers everything you need to add observability to your AI agents.

Overview

Risicare provides multiple ways to instrument your code:

Core SDK configuration and decorators

OpenAI, Anthropic, Cohere, and more

LangGraph, CrewAI, AutoGen integrations

Start simple and add depth as needed:

Set RISICARE_TRACING=true for automatic instrumentation of all LLM calls.

export RISICARE_API_KEY=rsk-xxx
export RISICARE_TRACING=true
python agent.py

Call risicare.init() for configuration control.

import risicare
risicare.init(environment="production")

Use @agent() to identify agent functions.

@agent(name="planner", role="orchestrator")
def plan(objective):
    pass

Group traces with @session or session_context().

with session_context(session_id=user_session):
    agent.run(query)

Track decision phases with @trace_think, @trace_decide, @trace_act.

@trace_think
def analyze(): pass
 
@trace_decide
def choose(): pass
 
@trace_act
def execute(): pass

Track messages with @trace_message, @trace_delegate.

@trace_message
def send_to_reviewer(msg): pass
 
@trace_delegate
def assign_subtask(task): pass

All init() options and environment variables

@agent, @session, @trace_* decorators