Installation
Requires Python 3.9 or later.
pip install respan-tracing
RespanTelemetry.
export RESPAN_API_KEY="your-api-key"
export RESPAN_BASE_URL="https://api.respan.ai/api" # optional, this is the default
Trace a workflow
from respan_tracing import RespanTelemetry, workflow, task
telemetry = RespanTelemetry(api_key="your-api-key")
@task(name="compute")
def compute(x: int, y: int):
return x + y
@workflow(name="math_workflow")
def math_workflow():
result = compute(2, 3)
return result
print(math_workflow()) # 5
@workflow decorator creates a root trace, and @task creates a child span nested inside it.
Trace an LLM call
Auto-instrumentation captures OpenAI (and other LLM provider) calls automatically.
from respan_tracing import RespanTelemetry, workflow, task
from openai import OpenAI
telemetry = RespanTelemetry(api_key="your-api-key")
client = OpenAI()
@task(name="generate_joke")
def generate_joke():
response = client.chat.completions.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Tell me a joke about tracing"}],
)
return response.choices[0].message.content
@workflow(name="joke_workflow")
def joke_workflow():
return generate_joke()
print(joke_workflow())
Class-based workflows
Use the method_name parameter to decorate a class and specify the entry point method.
from respan_tracing import RespanTelemetry, workflow, task
from openai import OpenAI
telemetry = RespanTelemetry(api_key="your-api-key")
client = OpenAI()
@workflow(name="joke_agent", method_name="run")
class JokeAgent:
@task(name="joke_creation")
def create_joke(self):
completion = client.chat.completions.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Tell me a joke about AI"}],
)
return completion.choices[0].message.content
def run(self):
return self.create_joke()
print(JokeAgent().run())
get_client() to attach customer identifiers, metadata, and other Respan-specific attributes.
from respan_tracing import RespanTelemetry, workflow, get_client
telemetry = RespanTelemetry(api_key="your-api-key")
@workflow(name="user_workflow")
def user_workflow(user_id: str):
client = get_client()
client.update_current_span(
respan_params={
"customer_identifier": user_id,
"metadata": {"source": "quickstart"},
}
)
return "done"
user_workflow("user-123")
View your traces
Open the Traces page in the Respan dashboard to see your trace tree.
Next steps
