respan-instrumentation-crewai

First-party Respan instrumentation for CrewAI. It subscribes to CrewAI's official lifecycle events and emits canonical workflow, task, agent, tool, and LLM spans directly. LLM spans include model, prompt/completion, and provider-reported token usage, which allows Respan to calculate cost for models in its pricing catalog.

Configuration

1. Install

pip install respan-instrumentation-crewai

The package requires CrewAI 1.10.1 or newer. That version introduced the provider-neutral LLM lifecycle and usage hooks required for complete tracing.

2. Set Environment Variables

Quickstart

3. Run Script

import os
from dotenv import load_dotenv

load_dotenv()

# Route OpenAI traffic through the Respan gateway (no separate OpenAI key needed)
os.environ["OPENAI_API_KEY"] = os.environ["RESPAN_API_KEY"]
os.environ["OPENAI_BASE_URL"] = os.getenv("RESPAN_BASE_URL", "https://api.respan.ai/api")

from respan import Respan
from respan_instrumentation_crewai import CrewAIInstrumentor

respan = Respan(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url=os.getenv("RESPAN_BASE_URL", "https://api.respan.ai/api"),
    instrumentations=[CrewAIInstrumentor()],
)

from crewai import Agent, Task, Crew

agent = Agent(
    role="Poet",
    goal="Write a short haiku about recursion in programming",
    backstory="You are a programmer who writes haikus.",
    llm="gpt-4o-mini",
)

task = Task(
    name="Write recursion haiku",
    description="Write a haiku about recursion in programming.",
    expected_output="A single haiku (3 lines: 5-7-5 syllables).",
    agent=agent,
)

crew = Crew(name="crewai_haiku", agents=[agent], tasks=[task])
result = crew.kickoff()
print(result.raw)

respan.flush()

4. View Dashboard

After running the script, traces appear on your Respan dashboard.

Further Reading

See the Respan example projects for runnable scripts.