smolagents (Hugging Face)

smolagents is a minimalist, open-source AI agent framework developed by Hugging Face. Muster captures smolagents runs through the SmolagentsInstrumentor.

Setup

1. Install dependencies

%pip install langfuse 'smolagents[telemetry]' opentelemetry-sdk opentelemetry-exporter-otlp openinference-instrumentation-smolagents

2. Configure credentials

import os

os.environ["LANGFUSE_PUBLIC_KEY"] = "pk-lf-..."
os.environ["LANGFUSE_SECRET_KEY"] = "sk-lf-..."
os.environ["LANGFUSE_BASE_URL"] = "https://app.getmuster.io"
os.environ["HF_TOKEN"] = "hf_..."

3. Initialize the client

from langfuse import get_client

langfuse = get_client()

if langfuse.auth_check():
    print("Muster client is authenticated and ready!")

4. Instrument smolagents

from openinference.instrumentation.smolagents import SmolagentsInstrumentor

SmolagentsInstrumentor().instrument()

After this, every CodeAgent / ToolCallingAgent run is captured as a trace with sub-spans for every tool call and LLM invocation.

Add user/session attributes

from langfuse import propagate_attributes

with propagate_attributes(user_id="user_123", session_id="sess_abc", tags=["smolagents"]):
    # run your smolagent
    ...

Troubleshooting

Enable debug logging: export LANGFUSE_DEBUG="True".
Call langfuse.flush() to ensure exports complete in short-lived scripts.