# 🔍 Fetch AI build failure summaries via the API

**URL:** <https://discuss.bitrise.io/t/fetch-ai-build-failure-summaries-via-the-api/25956>\
**Category:** Changelog\
**Tags:** api-update\
**Created:** [May 15, 2026, 3:18pm UTC](https://discuss.bitrise.io/t/fetch-ai-build-failure-summaries-via-the-api/25956 "2026-05-15T15:18:01Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![jeremyp](https://sea2.discourse-cdn.com/flex016/user_avatar/discuss.bitrise.io/jeremyp/32/5490_2.png) [@jeremyp](https://discuss.bitrise.io/u/jeremyp)\
**Post date:** [May 15, 2026, 3:18pm UTC](https://discuss.bitrise.io/t/fetch-ai-build-failure-summaries-via-the-api/25956/1 "2026-05-15T15:18:01Z")

</div>

When a build fails, Bitrise is able to generate an AI-powered summary that explains what went wrong and why. You can now retrieve that summary programmatically via the API — so you can bring the insight directly into the tools your team already uses, without anyone having to open the Bitrise UI.

Available on all plans that support the AI Build Summary feature.

* * *

### What’s New

**`GET /builds/{build-slug}/ai-summary`**

A new API endpoint that returns the AI-generated failure summary for a completed build. The endpoint is read-only and works with your existing API token.

Check out the build section of the [API documentation](https://docs.bitrise.io/en/bitrise-ci/api/api-reference.html) to find out how to implement it.

* * *

### The Problem

When a build fails, the information that explains _why_ is only available inside the Bitrise UI. For teams running high build volumes, that means context-switching: someone gets a Slack notification, navigates to the PR, opens Bitrise, reads the logs, and only then understands what broke. That loop is slow — and for on-call engineers or automated triage workflows, it’s a real bottleneck.

* * *

### What You Can Build With This

- **Slack notifications with failure context** — enrich your existing failure alerts with the AI summary so engineers get the “what broke and why” without leaving Slack
- **Bitrise workflow steps** — call the endpoint as a post-build step to route, label, or escalate failures based on the summary content
- **Internal dashboards** — aggregate and display failure summaries across builds to spot patterns
- **CI triage bots** — feed the summary into your own tooling or LLM pipeline for automated root-cause classification

* * *

### Getting Started

1. **Enable AI Build Summary** on your app — the endpoint returns 404 until this is active. **[AI Build Summary — Bitrise Docs](https://docs.bitrise.io/en/bitrise-ci/run-and-analyze-builds/build-data-and-troubleshooting/ai-build-summary.html)** _— how to enable the feature in Project Settings_
2. **Call the endpoint** after a build fails:

```auto
GET https://api.bitrise.io/v0.1/builds/{build-slug}/ai-summary

```

1. Use the response wherever it’s most useful — Slack, your dashboard, a workflow step.

* * *

Happy debugging 👋

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/flex016/uploads/bitrise/original/2X/d/d901cf2aad020f2dd11918afc89ec2089bb87b35.png) [@system](https://discuss.bitrise.io/u/system)\
**Post date:** [August 13, 2026, 3:18pm UTC](https://discuss.bitrise.io/t/fetch-ai-build-failure-summaries-via-the-api/25956/2 "2026-08-13T15:18:35Z")

</div>

This topic was automatically closed after 90 days. New replies are no longer allowed.