LangGraph | Respan Docs

LangGraph is a Python framework for building stateful, multi-step agent workflows as graphs. Nodes represent operations (LLM calls, tools, routing), and edges define the flow between them. Respan gives you full observability over every graph run, node, and LLM call — and gateway routing through the OpenAI-compatible Respan endpoint.

Set up Respan

Create an account at platform.respan.ai and grab an API key. For gateway, also add credits or a provider key.

Run npx @respan/cli setup to set up with your coding agent.

Example projects

Python examples

Tracing

Gateway

Setup

Install packages

$ pip install respan-ai openinference-instrumentation-langchain langgraph langchain-openai

Set environment variables

$ export OPENAI_API_KEY="YOUR_OPENAI_API_KEY"
$ export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"

OPENAI_API_KEY is used for LLM requests. RESPAN_API_KEY is used to export traces to Respan.

Initialize and run

1 from typing import TypedDict
2 from langgraph.graph import StateGraph, START, END
3 from langchain_openai import ChatOpenAI
4 from respan import Respan
5 from openinference.instrumentation.langchain import LangChainInstrumentor
6 
7 respan = Respan(instrumentations=[LangChainInstrumentor()])
8 
9 llm = ChatOpenAI(model="gpt-4.1-nano")
10 
11 class State(TypedDict):
12     topic: str
13     joke: str
14     review: str
15 
16 def generate_joke(state: State) -> dict:
17     response = llm.invoke(f"Tell me a joke about {state['topic']}")
18     return {"joke": response.content}
19 
20 def review_joke(state: State) -> dict:
21     response = llm.invoke(f"Rate this joke: {state['joke']}")
22     return {"review": response.content}
23 
24 graph = StateGraph(State)
25 graph.add_node("generate", generate_joke)
26 graph.add_node("review", review_joke)
27 graph.add_edge(START, "generate")
28 graph.add_edge("generate", "review")
29 graph.add_edge("review", END)
30 
31 app = graph.compile()
32 result = app.invoke({"topic": "AI tracing"})
33 print(result)
34 respan.flush()

View your trace

Open the Traces page to see the graph execution with node spans, LLM calls, and state transitions.

Configuration

Parameter	Type	Default	Description
`api_key`	`str \| None`	`None`	Falls back to `RESPAN_API_KEY` env var.
`base_url`	`str \| None`	`None`	Falls back to `RESPAN_BASE_URL` env var.
`instrumentations`	`list`	`[]`	Plugin instrumentations to activate (e.g. `LangChainInstrumentor()`).
`customer_identifier`	`str \| None`	`None`	Default customer identifier for all spans.
`metadata`	`dict \| None`	`None`	Default metadata attached to all spans.
`environment`	`str \| None`	`None`	Environment tag (e.g. `"production"`).

Attributes

In Respan()

Set defaults at initialization — these apply to all spans.

1 from respan import Respan
2 from openinference.instrumentation.langchain import LangChainInstrumentor
3 
4 respan = Respan(
5     instrumentations=[LangChainInstrumentor()],
6     customer_identifier="user_123",
7     metadata={"service": "graph-api", "version": "1.0.0"},
8 )

With propagate_attributes

Override per-request using a context scope.

1 from respan import Respan, propagate_attributes
2 from openinference.instrumentation.langchain import LangChainInstrumentor
3 
4 respan = Respan(instrumentations=[LangChainInstrumentor()])
5 
6 def handle_request(user_id: str, topic: str):
7     with propagate_attributes(
8         customer_identifier=user_id,
9         thread_identifier="conv_abc_123",
10         metadata={"plan": "pro"},
11     ):
12         result = app.invoke({"topic": topic})
13         print(result)

Attribute	Type	Description
`customer_identifier`	`str`	Identifies the end user in Respan analytics.
`thread_identifier`	`str`	Groups related messages into a conversation.
`metadata`	`dict`	Custom key-value pairs. Merged with default metadata.

Decorators (optional)

Decorators are not required. All LangGraph node executions and LLM calls are auto-traced by the instrumentor. Use @workflow and @task to add structure when you want to group related graph runs into a named workflow with nested tasks.

1 from respan import Respan, workflow, task
2 from openinference.instrumentation.langchain import LangChainInstrumentor
3 
4 respan = Respan(instrumentations=[LangChainInstrumentor()])
5 
6 @task(name="run_joke_graph")
7 def run_joke_graph(topic: str):
8     return app.invoke({"topic": topic})
9 
10 @workflow(name="joke_pipeline")
11 def pipeline(topic: str):
12     print(run_joke_graph(topic))
13 
14 pipeline("AI tracing")
15 respan.flush()