Skip to main content
GET
/
api
/
v1
/
guardrails
List Guardrails
curl --request GET \
  --url http://localhost:8000/api/v1/guardrails
{
  "agent_id": "<string>",
  "guardrail_name": "<string>",
  "severity": "<string>",
  "mode": "<string>",
  "timing": "<string>",
  "judge_model": "<string>",
  "description": "<string>",
  "judge_prompt": "<string>",
  "health": "<string>",
  "health_reason": "<string>",
  "registered_at": "<string>",
  "last_seen_at": "<string>",
  "recent_activity_24h": 123,
  "provider": "<string>"
}

Documentation Index

Fetch the complete documentation index at: https://docs.tracectrl.ai/llms.txt

Use this file to discover all available pages before exploring further.

Returns every guardrail in the registry. Optionally filter to a single agent.

​
Query Parameters

​
agent_id
string
Filter to guardrails registered for a specific agent.

​
Response

Array of GuardrailRegistration objects:
​
agent_id
string
Agent the guardrail is attached to
​
guardrail_name
string
Guardrail identifier
​
severity
string
low, medium, high, or critical
​
mode
string
monitoring or blocking
​
timing
string
pre_input or post_output
​
judge_model
string
Model used for evaluation
​
description
string
Human-readable description
​
judge_prompt
string
Custom judge prompt (may be empty)
​
health
string
active, error, or disabled
​
health_reason
string
Explanation when health is not active
​
registered_at
string
ISO-8601 timestamp
​
last_seen_at
string
ISO-8601 timestamp
​
recent_activity_24h
integer
Invocations in the last 24h
​
provider
string
judge_llm or protector_plus