Add Observability to Your AI Agent in 3 Lines

Drop-in integrations for LangChain, OpenAI Agents, Claude, and CrewAI. No manual instrumentation required.

The Problem

You’re using a framework to build your agent. You want to see every chain step, tool call, and LLM interaction — without wrapping each one in custom logging. You need observability that plugs into your existing code, not a rewrite.

The Solution

pip install axonpush[langchain]       # LangChain / LangGraph
pip install axonpush[openai-agents]   # OpenAI Agents SDK
pip install axonpush[anthropic]       # Claude / Anthropic
pip install axonpush[crewai]          # CrewAI
pip install axonpush[deepagents]     # LangChain Deep Agents

LangChain / LangGraph

from axonpush import AxonPush
from axonpush.integrations.langchain import AxonPushCallbackHandler

client = AxonPush(api_key="ak_...", tenant_id="1", base_url="https://api.axonpush.xyz")
handler = AxonPushCallbackHandler(client, channel_id=1, agent_id="my-agent")

chain.invoke({"input": "research AI frameworks"}, config={"callbacks": [handler]})

OpenAI Agents SDK

from axonpush import AsyncAxonPush
from axonpush.integrations.openai_agents import AxonPushRunHooks

client = AsyncAxonPush(api_key="ak_...", tenant_id="1", base_url="https://api.axonpush.xyz")
hooks = AxonPushRunHooks(client, channel_id=1)

result = await Runner.run(agent, input="research AI frameworks", hooks=hooks)

Claude / Anthropic

from axonpush import AxonPush
from axonpush.integrations.anthropic import AxonPushAnthropicTracer

client = AxonPush(api_key="ak_...", tenant_id="1", base_url="https://api.axonpush.xyz")
tracer = AxonPushAnthropicTracer(client, channel_id=1)

response = tracer.create_message(
    anthropic_client,
    model="claude-sonnet-4-20250514",
    messages=[{"role": "user", "content": "Research AI frameworks"}],
)

CrewAI

from axonpush import AxonPush
from axonpush.integrations.crewai import AxonPushCrewCallbacks

client = AxonPush(api_key="ak_...", tenant_id="1", base_url="https://api.axonpush.xyz")
callbacks = AxonPushCrewCallbacks(client, channel_id=1)

callbacks.on_crew_start()
result = Crew(
    agents=[researcher, writer],
    tasks=[research_task, write_task],
    step_callback=callbacks.on_step,
    task_callback=callbacks.on_task_complete,
).kickoff()
callbacks.on_crew_end(result)

LangChain Deep Agents

from deepagents import create_deep_agent
from axonpush import AxonPush
from axonpush.integrations.deepagents import AxonPushDeepAgentHandler

client = AxonPush(api_key="ak_...", tenant_id="1", base_url="https://api.axonpush.xyz")
handler = AxonPushDeepAgentHandler(client, channel_id=1, agent_id="deep-agent")

agent = create_deep_agent(tools=[], system_prompt="You are a helpful assistant.")
agent.invoke(
    {"messages": [{"role": "user", "content": "Research AI frameworks"}]},
    config={"callbacks": [handler]},
)

Events Published by Each Integration

Framework	Events
LangChain	chain.start, chain.end, chain.error, llm.start, llm.end, tool.<name>.start, tool.end
OpenAI Agents	agent.run.start, agent.run.end, tool.<name>.start, tool.<name>.end, agent.handoff
Anthropic	conversation.turn, tool.<name>.start, agent.response, tool.result
CrewAI	crew.start, crew.end, agent.step, tool.<name>.start, tool.<name>.end, task.complete
Deep Agents	chain.start/end, llm.start/end, planning.update/complete, subagent.spawn/complete, filesystem.read/write, sandbox.execute

Go Deeper

Cross-service trace correlation

Pass the same trace_id across services to build a unified trace:

handler = AxonPushCallbackHandler(client, channel_id=1, trace_id="tr_shared_123")
hooks = AxonPushRunHooks(client, channel_id=2, trace_id="tr_shared_123")

Both services’ events appear in the same trace when you call client.traces.get_events("tr_shared_123").