What data we store
- text (Skyline page content)
- images uploaded by the user within the app
- AI prompt written by the user excluding images
- AI response
- accountId and sitename
Where we store the data
| Data type | Where do we store it |
|---|---|
| Skyline page text | Confluence macro data (customer’s Confluence instance) |
| images uploaded by the user | Confluence page attachments (customer’s Confluence instance) |
| AI prompt | Supabase |
| AI response | Supabase |
| accountId and sitename (for payments processing, AI Credits balance and audit log of AI requests) | Supabase |
Payments processing
We use Stripe for processing AI Credits payments. We don’t send any identifiers to Stripe - we store payment session details in our DB (Supabase) and match them with the checkout session.AI processing details
We send the following data during Skyline page generation (AI generation screen):- user prompt
- image attached by the user to the prompt
- user prompt
- image attached by the user to the prompt
- Skyline page content (text added/generated in the app, urls of images uploaded in the app and urls of the images/videos used as background/image element source)
AI Governance
The AI Audit log access is available upon request (Enterprise AI Credits plan).Data access
The app’s purpose is to add new visually formatted content to the Confluence page. The only read of the existing Confluence content is checking the current page status (draft/published) to be able to upload attachments to the page (user-uploaded images) when the page is in a draft state. Besides that, the app doesn’t read classic Confluence pages’ content.Third-Party Services
| Service | Why we need it |
|---|---|
| Supabase | DB for AI prompt, AI response, AI Credits and payment details |
| Vercel | App Hosting, Analytics, Backend Logs |
| Google Gemini API | Access to Gemini 3 Pro Preview and Gemini 2.5 Pro AI models |
| OpenAI API | Access to GPT-5.1 and o4-mini-deep-research AI models |
| Posthog | App analytics |
| AWS S3 | Storage of our image and video libraries |
| Stripe | Payments processing |
| Unsplash | To enable free stock image search within the app |
Does the app support Data Residency?
| Data type | Data residency |
|---|---|
| Skyline page text and images | Data Residency is supported out of the box by Atlassian Cloud. The data is stored on the customer’s Confluence instance, so it remains in the same location as Confluence data. |
| AI prompt, AI response, accountId, sitename | Data Residency not supported - data is stored in eu-west-1 region (Dublin, Ireland). |
Why do we need specific permissions
read:connect-confluence - to be able to check the current page status (draft/published), so we can upload attachment to the page (image added by the user) while it’s still in a draft state
write:connect-confluence - to be able to upload images as page attachments