Guardrails (tracing) | Respan Docs

Guardrails AI is a framework for adding structural, type, and quality guarantees to LLM outputs. It validates, corrects, and structures LLM responses to ensure they meet your requirements. Respan gives you full observability over every guard validation, re-ask loop, validator execution, 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.

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

Use Respan Gateway

See Guardrails AI gateway setup to route this integration through the Respan gateway.

Example projects

Example projects root

Setup

Install packages

pip install respan-ai respan-instrumentation-guardrails guardrails-ai

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

import os
from typing import Literal
from guardrails import Guard
from pydantic import BaseModel, Field
from respan import Respan
from respan_instrumentation_guardrails import GuardrailsInstrumentor
class SupportReply(BaseModel):
    answer: str = Field(description="Customer-facing answer")
    priority: Literal["low", "medium", "high"] = Field(description="Ticket priority")
respan = Respan(
    api_key=os.environ["RESPAN_API_KEY"],
    instrumentations=[GuardrailsInstrumentor()],
)
guard = Guard.for_pydantic(output_class=SupportReply)
result = guard(
    model="gpt-4o-mini",
    messages=[{
        "role": "user",
        "content": "Return JSON with answer and priority for a delayed shipment."
    }],
)
print(result.validated_output)

View your trace

Open the Traces page to see your Guardrails workflow with validation passes, re-ask loops, and LLM calls.

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. `GuardrailsInstrumentor()`).
`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.

from respan import Respan
from respan_instrumentation_guardrails import GuardrailsInstrumentor
respan = Respan(
    instrumentations=[GuardrailsInstrumentor()],
    customer_identifier="user_123",
    metadata={"service": "guardrails-api", "version": "1.0.0"},
)

With propagate_attributes

Override per-request using a context scope.

from respan import Respan, propagate_attributes
from respan_instrumentation_guardrails import GuardrailsInstrumentor
respan = Respan(instrumentations=[GuardrailsInstrumentor()])
def handle_request(user_id: str, prompt: str):
    with propagate_attributes(
        customer_identifier=user_id,
        thread_identifier="conv_abc_123",
        metadata={"plan": "pro"},
    ):
        result = guard(model="gpt-4o-mini", messages=[{"role": "user", "content": prompt}])
        print(result.validated_output)

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.