Kitefold
Privacy

Privacy policy

Last updated 12 November 2026

In short: Kitefold's apps run on Atlassian's own platform and send nothing outside Atlassian. The All-Spaces Task Report reads your tasks as the person viewing them and stores only the settings for each weekly digest, plus one site-wide note for its review request. Link Health & Cleanup keeps an index of the links on your site (ids, titles, link targets, file names and sizes, never page text) in Atlassian's storage for the app, and checks everything it shows as the person viewing it. Query & Bulk Actions searches and changes Confluence as the person using it, in their browser, and stores only saved searches, one setting and usage counts; its optional plain-English search uses an AI model hosted by Atlassian that never sees page content. This website uses no cookies and no tracking.

1. Who we are

Kitefold is a sole trader business in the United Kingdom. In this policy, “Kitefold”, “we” and “us” mean Kitefold.

Contact: hello@kitefold.app (support questions: support@kitefold.app)

This policy covers Kitefold's Confluence apps (All-Spaces Task Report, Link Health & Cleanup and Query & Bulk Actions, all for Confluence Cloud), our support email and this website (kitefold.app).

2. The All-Spaces Task Report app

Where it runs

The app is built on Atlassian Forge. It runs on Atlassian's cloud platform, inside the same Atlassian cloud as your Confluence site. Kitefold has no servers of its own, and the app makes no calls to any service outside Atlassian. Your organisation's Confluence content stays under your organisation's control and Atlassian's terms; see Atlassian's privacy policy.

What it reads

To show the report, the app reads from Confluence: tasks (their text, status, due date and assignee), the titles and links of the pages and blog posts they're on, and space names and keys. People are shown by their Atlassian account, using Confluence's own name and picture.

The report reads all of this as the person viewing it, from their browser through Atlassian, so Confluence only returns what that person is allowed to see. It's used only to display the report on screen and is not saved.

What it changes

  • Ticking a task in the report changes that task's status (done or not done), as the person who ticked it. Confluence refuses the change if they can't edit the page.
  • The weekly digest, once someone turns it on, creates a new Confluence page each week under the page that holds the digest. That page lists each person with tasks overdue or due in the next 7 days, with an @mention and two numbers. It contains no task text and no page titles. Confluence notifies the people mentioned in the usual way. Anyone who can view the digest's page can see the names and numbers on it.

How the digest counts tasks

The digest runs on a schedule when nobody is viewing, so it uses Forge's ability to act on behalf of a user. It reads tasks and spaces first as the person who last turned the digest on or saved its settings, then as each person it lists, so every number is what that person could see for themselves. This is only used for reading tasks and spaces. The digest page itself is created by the app.

What it stores

The app stores no task data. For each weekly digest, it keeps these settings in Atlassian's Forge storage:

  • the page the digest belongs to, the spaces it counts and the day it posts
  • whether it's on
  • the Atlassian account ID of the person who last turned it on or saved its settings (so the digest can read tasks as them)
  • the title, link and time of the last digest posted, and the last error, if any
  • this week's try count: the day the digest last tried to post and how many times it tried that day (at most twice, so a failed post is retried once and never more)

It also keeps one note for the whole site: when the first scheduled digest posted, and whether someone has dismissed the short request for a Marketplace review that the digest shows to page editors. It holds those two times only: no names and no account IDs.

Pressing Turn off deletes a digest's settings. When the app is uninstalled, Atlassian deletes the app's stored data under its own retention rules for Forge apps (it can be restored if the app is reinstalled within 21 days).

Data residency

All the data the app stores is the digest settings and the site-wide review note listed above. They're kept only in Atlassian's Forge hosted storage, so they follow the data residency location your organisation has chosen for its Atlassian apps, and move with it if that location changes.

  • In scope for data residency: the weekly digest settings above, including the Atlassian account ID of the person who set the digest up, and the site-wide review note.
  • Out of scope: nothing. The app stores no other data. Tasks, page titles and people's names are read from Confluence when a report or digest runs and are never stored.

Logs

Like all Forge apps, the app writes technical logs that Atlassian keeps for a limited time and that Kitefold can read to fix problems. Our logs record things like how many people a digest counted, error codes, and Confluence's rate-limit readings (how much of the app's hourly allowance is left). They are designed not to include task text or page content.

What the app doesn't do

  • It doesn't send your data outside Atlassian.
  • It doesn't use analytics, tracking or advertising.
  • It doesn't send your content to AI services.
  • It doesn't sell or share your data.

Where it runs

Like all Kitefold apps, it's built on Atlassian Forge and runs inside the same Atlassian cloud as your Confluence site. Kitefold has no servers of its own, and the app makes no calls to any service outside Atlassian.

What it reads

To build its index, the app reads Confluence as the app itself (not as any person), in the spaces it's allowed to read: each page's and blog post's title, space, status and short link, and the links in its content; the list of spaces; and each attachment's file name, size and the page it's on, plus which image a page uses as its header. It reads page content only to find the links in it; the text itself is not kept. Spaces an admin leaves out on the Settings tab aren't read at all. When a page or blog post changes, Confluence tells the app, and it reads that one item again.

Before showing anything, such as a list row, a page title or a “Linked from” count, the app checks with Confluence as the person viewing it, and leaves out anything they can't open. The Stale pages list is a Confluence search made as that person, and isn't stored.

What it changes

  • Archiving pages and moving files to the trash, only when an admin selects them and confirms on the admin page. Both run as that admin, so Confluence refuses anything they aren't allowed to change, and both can be undone in Confluence. The app never deletes anything permanently and never edits page content.

What it stores

In Atlassian's Forge storage for the app:

  • for each page and blog post: its id, title, space, status, short-link code and the targets of its links (page titles, ids, file names and link addresses), and for blog posts the date they were posted
  • for each file: its id, file name, size, status and the page it's on
  • for each space: its id, key, name and homepage
  • which pages link to each page (ids only), and the latest report: totals and the lists shown on the admin page
  • settings chosen by admins: old server addresses, spaces left out, pages marked “not an orphan”, the stale pages age and the Attachments macro choice
  • the ids of pages and files archived or moved to the trash through the app, until the next full index
  • indexing progress and daily usage counters (numbers only)
  • two times for the one-time review request on the admin page: when an admin first saw a finished report, and when someone dismissed the request

It stores no page text, no people's names and no Atlassian account IDs. The index is rebuilt regularly, and records for deleted pages and files are removed. When the app is uninstalled, Atlassian deletes the app's stored data under its own retention rules for Forge apps (it can be restored if the app is reinstalled within 21 days).

Everything the app stores is listed above and is kept only in Atlassian's Forge hosted storage, so it follows the data residency location your organisation has chosen for its Atlassian apps, and moves with it if that location changes.

  • In scope for data residency: the index, the latest report, the settings and the review request times listed above.
  • Out of scope: nothing. The app stores no other data.

Logs

The app's technical logs, kept by Atlassian for a limited time, record things like counts, page and file ids, space keys, error codes and the old server address an admin entered. They are designed not to include page text.

What the app doesn't do

  • It doesn't send your data outside Atlassian, and doesn't check links to other websites.
  • It doesn't use analytics, tracking or advertising.
  • It doesn't send your content to AI services.
  • It doesn't sell or share your data.

4. The Query & Bulk Actions app

Where it runs

Like all Kitefold apps, it's built on Atlassian Forge and runs inside the same Atlassian cloud as your Confluence site. Kitefold has no servers of its own, and the app makes no calls to any service outside Atlassian.

What it reads

The app reads Confluence as the person using it, in their browser: the search results they ask for (titles, types, spaces, dates, people, labels, versions, parent pages and file sizes, as they choose), space and label names for the pickers, and people's names. It never sees more than that person can. Search results are shown on the page and, if the person exports them, saved to a file on their own computer; they're never stored by the app.

What it changes

  • Only what the person chooses to change, after a preview and a confirmation, and as that person, so Confluence refuses anything they aren't allowed to change: adding and removing labels (including renaming or merging a label across the site), watching and stopping watching pages (for themselves, or for someone else if they're a Confluence admin or space admin), and archiving pages. It never deletes anything and never edits page content.

What it stores

In Atlassian's Forge storage for the app:

  • saved searches: the name, the search text (CQL), columns, sort and number of rows, whether it's shared, dates, and the Atlassian account ID of the person who saved it and of a Confluence admin who last edited it
  • one setting: whether bulk actions are shown on the search page
  • usage counts with no personal data: an hourly record of how much the site has asked of Confluence (kept 48 hours), a weekly count of how often Confluence asked the app to slow down, and a monthly count of plain-English searches with the AI model used and its token totals
  • for the current month only, the Atlassian account IDs of people who have used “Ask in plain English”, so the site's allowance can grow with them; they expire automatically on the 3rd of the next month

It stores no search results, page content, page titles or people's names. In the person's own browser it keeps a few minutes of a Query results macro's results (so the page loads quickly), their column and row choices, and a note of a bulk run in progress (so it can say if the run was cut short). When the app is uninstalled, Atlassian deletes the app's stored data under its own retention rules for Forge apps (it can be restored if the app is reinstalled within 21 days).

Ask in plain English (AI)

When someone uses “Ask in plain English”, their request, today's date, the list of search fields the app allows, the names of spaces they can see and the names of up to 20 labels that share a word with the request are sent to an AI model hosted by Atlassian inside its own platform (Forge LLMs). Page content, page titles and people are never sent. The AI writes a search that the person sees and can change before it runs. Kitefold doesn't store or log what anyone types, and nothing is sent to an AI service outside Atlassian. The feature is optional; the rest of the app works without it.

Atlassian has confirmed to us how Forge LLMs handle these requests:

  • requests and answers are not stored by Atlassian or by the model provider, and are used only to answer that request;
  • they are never used to train or improve AI models, and aren't available to other customers;
  • the models run on Amazon Bedrock, which Atlassian lists among its sub-processors;
  • if Atlassian's moderation blocks a request, nothing is recorded about the person who made it.

Data residency

Everything the app stores is listed above and is kept only in Atlassian's Forge hosted storage, so it follows the data residency location your organisation has chosen for its Atlassian apps.

  • In scope for data residency: saved searches, the setting and the usage counts listed above.
  • Out of scope: nothing stored. Plain-English requests are processed by Atlassian's hosted AI and not stored by the app or by Atlassian.

Where AI requests are processed is separate from where data is stored. For now, Atlassian processes Forge LLM requests in a single region, whatever your data residency location; this fits Atlassian's data residency policy, which covers stored data, because nothing is stored. Atlassian plans to let organisations choose where AI requests are processed (for example, only within the EU), and plain-English requests will follow that setting once it's available.

Logs

The app's technical logs, kept by Atlassian for a limited time, record counts, the AI model name, token totals, error codes and Confluence's usage-limit readings. They never include what anyone types, search results or page content.

What the app doesn't do

  • It doesn't send your data outside Atlassian.
  • It doesn't use analytics, tracking or advertising.
  • It doesn't send page content to any AI service.
  • It doesn't sell or share your data.

5. Licences and purchases

The apps are sold through the Atlassian Marketplace, and Atlassian handles all payments. We never see card or bank details. Atlassian shares licence information with us, which can include your organisation's name, your Confluence site address, and the names and email addresses of your technical and billing contacts. We use it only to manage licences, provide support and send important notices about the app. We keep it while your licence is active and afterwards for as long as the law requires us to keep business records (usually up to 6 years).

6. Support emails

If you email us, we receive your name, email address and whatever you send, and use it only to help you (our support hours and response times). We keep support emails for up to 2 years after the conversation ends, unless they're part of a business record we must keep for longer. Please don't send passwords, API tokens or confidential content.

7. This website

kitefold.app sets no cookies and uses no analytics, tracking, adverts, embedded videos, external fonts or third-party scripts. It's hosted on GitHub Pages, run by GitHub, Inc. Like any web host, GitHub may record visitors' IP addresses in its server logs for security and to run the service; see GitHub's privacy statement. We don't receive or use those logs.

The free Confluence link checker runs entirely in your browser. Whatever you paste into it stays on your device: it isn't uploaded, stored or sent anywhere, and the page makes no network requests of its own.

8. Why we're allowed to use data

Under UK data protection law (the UK GDPR and the Data Protection Act 2018), we rely on:

  • Legitimate interests: to provide the app's features your organisation chose to use, to fix problems, and to keep this website working and secure.
  • Contract: to manage licences and give support.
  • Legal obligation: to keep business and tax records.

9. Who else is involved

  • Atlassian hosts and runs the apps, handles Marketplace purchases and processes your Confluence data under its own terms. For Query & Bulk Actions' plain-English search, Atlassian also runs the AI model (Forge LLMs) on Amazon Bedrock, one of Atlassian's listed sub-processors.
  • Hostinger provides our email service, so support emails are stored with them.
  • GitHub hosts this website.

These providers may process data outside the UK. They use their own legal safeguards for international transfers. We don't share personal data with anyone else unless the law requires it.

10. Your rights

You have the right to ask for a copy of your personal data, and to ask us to correct it, delete it, restrict how we use it or stop using it. Email hello@kitefold.app and we'll reply within one month.

Your Confluence content belongs to your organisation, so for data inside your Confluence site (such as tasks, digest pages, Link Health's index or Query & Bulk Actions' saved searches), please contact your organisation's Confluence admin first.

If you're unhappy with how we've handled your data, you can complain to the Information Commissioner's Office (ICO) at ico.org.uk/make-a-complaint or on 0303 123 1113. We'd appreciate the chance to put things right first.

11. Changes to this policy

If an app or this website starts using data differently, we'll update this page and the date at the top before the change takes effect. Significant changes will also be noted in the app's Marketplace listing.