Set up the SDK

1
{
2
  "mcpServers": {
3
    "respan-docs": {
4
      "url": "https://mcp.respan.ai/mcp/docs"
5
    }
6
  }
7
}

The Respan tracing SDK provides decorators and wrappers to instrument your code. This page covers SDK setup for each framework, decorators for structuring traces, and advanced features.

1. Set your API key

$
export RESPAN_API_KEY="your-api-key"

For the fastest setup, use the CLI wizard instead: npx @respan/cli setup

2. Auto-instrumented SDKs

These LLM SDKs can be traced automatically — just install Respan and initialize. No instrumentor imports needed.

$
pip install respan-ai

1
from respan import Respan
2
Respan()
3
# All supported LLM calls are now auto-traced

3. Agent frameworks (explicit instrumentor)

Agent frameworks require an explicit instrumentor because they capture higher-level spans (agent runs, handoffs, tool calls) beyond raw LLM calls.

$
pip install respan-ai respan-instrumentation-openai-agents openai-agents

1
from agents import Agent, Runner
2
from respan import Respan
3
from respan_instrumentation_openai_agents import OpenAIAgentsInstrumentor
4
5
Respan(instrumentations=[OpenAIAgentsInstrumentor()])
6
7
agent = Agent(name="Assistant", instructions="Reply in one sentence.")
8
result = Runner.run_sync(agent, "Explain tracing.")
9
print(result.final_output)

Disable auto-instrumentation

By default, Respan auto-instruments all supported providers. You can disable specific ones:

1
from respan import Respan
2
3
# Disable specific providers
4
Respan(
5
    block_instruments={"bedrock", "vertexai"}
6
)

In Python, you can also enable only specific providers instead of disabling:

1
from respan import Respan
2
3
# Only instrument OpenAI and Anthropic
4
Respan(
5
    instruments={"openai", "anthropic"}
6
)

Add custom instrumentors

Agent frameworks and custom integrations require explicit instrumentors. Pass them via the instrumentations parameter:

1
from respan import Respan
2
from respan_instrumentation_openai_agents import OpenAIAgentsInstrumentor
3
4
Respan(instrumentations=[OpenAIAgentsInstrumentor()])

Next steps

• Customize traces — Use decorators to create nested, hierarchical traces
• Enrich span attributes — Attach metadata, customer identifiers, and other properties
• Manual ingestion — Send spans via OTLP or JSON API