---
title: "Ai Feedback Summary"
method: GET
path: "/feedback/ai/summary"
tags: ["feedback"]
---

# Ai Feedback Summary

`GET /feedback/ai/summary`

Admin readout: per-surface up/down counts and the latest comments/corrections.

Team rows are tagged, not dropped. The ratio is the thing team clicks
distort — six of us thumbs-downing a bad ChaCha reply moves a surface's
score in a way six users would not — so the aggregate counts only real
users, and `team` carries our own counts beside it. But a team member's
written correction is often the most useful row on the page, so every
comment still comes back, marked `is_team` for whoever is reading.

This endpoint never joined users at all, which is why it had no exclusion
to apply either way (2026-09-01 admin audit).

## Query parameters

- `days` integer
- `segment` string

## Headers

- `authorization` string, nullable

## Response `200`

Successful Response

- unknown

## Other responses

- `422` — Validation Error

## Changes

- **2026-09-02** `45cbba31aa7d` — 1 info
  - added the new optional `query` request parameter `segment`

[Change history](https://skmtc.dev/fluentea/apis/aurqa-language-learning-platform-refactored/changes/feedback/ai/summary/get.md)

---

[API](https://skmtc.dev/fluentea/apis/aurqa-language-learning-platform-refactored.md) · [All operations](https://skmtc.dev/fluentea/apis/aurqa-language-learning-platform-refactored/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/fluentea/aurqa-language-learning-platform-refactored/revisions/aea5e6e7d830/schema)