---
title: "Get System Guardrail Trigger Log"
method: GET
path: "/v1/data_extract/system_guardrail_trigger_log"
tags: ["Data Extract"]
---

# Get System Guardrail Trigger Log

`GET /v1/data_extract/system_guardrail_trigger_log`

## Request

- Base URL: `https://api.capacity.com` (the document declares no server; this is the origin it was published from)
- URL: `https://api.capacity.com/v1/data_extract/system_guardrail_trigger_log`
- Auth: one of:
  - API key in header `x-capacity-id`
  - HTTP bearer

Return one row per system guardrail trigger event, including the interface, transcript and audio references, the categories that fired, their scores, the model reasoning, the generation outcome, the applied tiers, the triggering text, the filter that fired, the action taken, and the matched phrase. Results are paginated and ordered by created_at_utc DESC.

## Query parameters

- `page` integer
- `limit` integer
- `start_datetime` string, required
- `end_datetime` string, required
- `interface` string

## Response `200`

A custom data array of System Guardrail Trigger Log objects

- object
  - `body` object: Json body data in the request
  - `limit` integer: Number of results returned
  - `page` integer: Page of results returned
  - `parameters` object: Query parameters in the request
  - `result` SystemGuardrailTriggerLogResponseSchema
    - `data` SystemGuardrailTriggerLogSchema[]
      - `created_at_utc` string, date-time: UTC timestamp when the guardrail trigger occurred
      - `interface_name` string: Name of the interface associated with the trigger
      - `audio_log_url` string: URL to download the audio of the interaction
      - `categories_fired` object: Categories that fired for the trigger
      - `category_scores` object: Scores associated with each fired category
      - `reasoning` string: Model reasoning for the guardrail trigger
      - `generation_outcome` string: Outcome of the generation after the guardrail evaluation
      - `applied_tiers` object: Tiers applied as a result of the trigger
      - `triggering_text` string: Text that triggered the guardrail
      - `filter_name` string, nullable: Name of the filter that fired
      - `action_taken` string, nullable: Action taken as a result of the filter firing
      - `matched_phrase` string, nullable: Phrase matched by the filter
    - `_meta` object

## Other responses

- `400`: The request contains malformed data or violates data integrity

---

[API](https://skmtc.dev/capacity/apis/capacity-api-docs.md) - [All operations](https://skmtc.dev/capacity/apis/capacity-api-docs/llms.txt) - [OpenAPI document](https://skmtc.dev/capacity/apis/capacity-api-docs/revisions/74c85bb61004?raw)