---
title: Cloud Assist FAQ
description: Frequently asked questions about Capawesome Cloud Assist and Ask AI — privacy, pricing, permissions, caching, and supported project types.
faq: true
---

# FAQ

## General

### What is Capawesome Cloud Assist?

[Cloud Assist](./index.md) is the AI surface inside the [Capawesome Cloud](/){:target="_blank"} Console. It groups all AI-powered features under a single name, starting with [Ask AI](./ask-ai.md) for failed builds and deployments.

### What's the difference between Cloud Assist and Ask AI?

**Cloud Assist** is the umbrella product. **Ask AI** is the first feature shipping under it. More features will be added over time.

### Is Cloud Assist available on all plans?

Cloud Assist is included free in every [Capawesome Cloud plan](https://capawesome.io/pricing/){:target="_blank"} that includes build minutes. Plans without build minutes don't include Cloud Assist, since the feature operates on build and deployment logs.

### Does Cloud Assist work for non-Capacitor projects?

Yes. Cloud Assist works across all project types supported by Capawesome Cloud — Capacitor, Cordova, native iOS, and native Android.

## Ask AI

### What data is sent to the AI provider?

Only the build logs of the specific failed job. No source code, no environment variables, no project metadata, and no other jobs are sent.

### Is my data used to train AI models?

No. Our AI provider does not retain or train on the data sent through Cloud Assist.

### Who can use Ask AI?

Only members of the organization with **write access to the app** can trigger Ask AI. This mirrors the permissions that already control builds and deployments. See [Roles and Permissions](../organizations/memberships.md#roles-and-permissions) for details.

### Are results cached?

Yes. The first teammate to click **Ask AI** on a failed job triggers the run, and the result is cached against that job and shared across the organization. Every other member with permission opens the dialog and sees the same answer instantly.

### Why is Ask AI marked as experimental?

Ask AI is shipping as a beta. Quality, latency, and UX will continue to iterate based on real usage and feedback. The "New" badge in the Console signals this.

### Can the AI's answer be wrong?

Yes. Like any AI-generated content, the diagnosis can be incorrect or incomplete. Always verify the suggested fix before applying it.

### Why is the output capped at around 400 words?

To keep the result a summary that's quick to read, copy, and paste — rather than a wall of text you'd have to re-read the logs to verify.

### Can I regenerate the answer or ask follow-up questions?

Not yet. The current version returns a single response. Regenerate-with-context and follow-up questions are on the roadmap.

### Does Ask AI work on succeeded builds?

No. The **Ask AI** button only appears on failed builds and deployments.