Security
The Aldorock apps for Jira, Jira Service Management and Confluence, by Aldorock AB (org. no. 559607-5985, Gammelgårdsvägen 58, 112 64 Stockholm, Sweden). Version 2, 18 November 2026. What each app stores is listed in the privacy policy. This page says how it is protected.
In short
Every app is built on Atlassian Forge. It runs on Atlassian's cloud as part of your Atlassian site, keeps its data only in Forge storage and makes no network calls to any host outside Atlassian. Aldorock AB runs no servers for these apps, keeps no copy of your data and never asks for passwords or API tokens.
Hosting and data location
- The apps run as Forge functions on Atlassian's infrastructure. None of them uses Forge Remote, web triggers or a backend of its own.
- App data is kept only in Forge hosted storage, which Atlassian encrypts at rest and which follows your site's data residency. Diagram Macros for Confluence stores nothing of its own: each diagram's source stays in the macro on the page and the rendered image is saved as a page attachment.
- Backups of Forge storage are made and encrypted by Atlassian. We keep no backups of customer data ourselves.
- The apps write only to Forge's own app logs, inside Atlassian. A log entry can hold an Atlassian account ID, a work item, request or page key, and an error message from Jira or Confluence, which may quote a title or summary. Logs never hold credentials, tokens or passwords.
- The apps use no AI or machine learning models, and no Atlassian data is sent to any AI service.
Access control
Each app asks Atlassian only for the scopes its features need, and Forge enforces them. An app's storage can be read only by that app running in your site. Aldorock AB staff cannot read your site's Forge storage.
Q&A for Confluence, Tree Copy for Confluence, Label Tools for Confluence and Templates for Jira make every Jira or Confluence call as the current user, so the product's own permissions apply. The other eight apps also make some calls as the app itself, and the table below says when. Where such a call serves a user, the app first checks that the user may see the data: through the Jira or Confluence permissions API, or for Portal Fields through the request's own reporter, participants and organizations.
Every input to an app's backend is checked before use: type, length and allowed values. IDs and keys are checked against the current context and the user's permissions. Text that people enter is shown through UI Kit or React, which escape it, and is never inserted as raw HTML. Diagram Macros renders Mermaid in strict mode with HTML labels off, and Markdown with raw HTML disabled and links limited to web, mail and relative addresses.
Scopes per app
| App | Jira or Confluence calls run as | Scopes |
|---|---|---|
| Templates for Jira | The user | read:jira-work write:jira-work read:jira-user storage:app |
| Q&A for Confluence | The user | read:confluence-user read:confluence-groups read:space:confluence read:page:confluence write:page:confluence storage:app |
| Epic Totals for Jira | The user, or the app after a permission check | read:jira-work write:jira-work read:jira-user storage:app |
| Follow-ups for Jira | The user, or the app after a permission check | read:jira-work write:jira-work read:jira-user storage:app |
| Tree Copy for Confluence | The user | read:confluence-user read:space:confluence read:page:confluence read:hierarchical-content:confluence read:content-details:confluence read:content.restriction:confluence read:content.property:confluence write:page:confluence read:label:confluence write:label:confluence read:attachment:confluence write:attachment:confluence read:confluence-content.all write:confluence-content search:confluence storage:app |
| Portal Fields for JSM | The user on the settings page. The portal panel reads the request as the app, after checking that the viewer is its reporter, a participant or in an organization it is shared with | read:servicedesk-request read:jira-work read:jira-user storage:app |
| Worklog Invoicing for Jira | The user, or the app after a permission check | read:jira-work write:jira-work read:jira-user storage:app |
| Read Acknowledgments for Confluence | The user. The app only when a page is deleted, to close that page's requests | read:confluence-user read:confluence-groups read:space:confluence read:page:confluence write:comment:confluence search:confluence read:content-details:confluence storage:app |
| Repeating Work for Jira | The user, or the app after a permission check | read:jira-work write:jira-work read:jira-user storage:app |
| Diagram Macros for Confluence | The user, and saving an image only after an edit permission check. The app only to list a page's attachments during a PDF or Word export | read:page:confluence read:attachment:confluence write:attachment:confluence delete:attachment:confluence write:confluence-file read:confluence-content.all |
| Default Field Values for Jira | The user, or the app after a permission check | read:jira-work write:jira-work read:jira-user storage:app |
| Label Tools for Confluence | The user | read:confluence-user read:space:confluence read:page:confluence read:hierarchical-content:confluence read:label:confluence write:label:confluence read:confluence-content.all write:confluence-content search:confluence storage:app |
Encryption
- In transit: the apps open no connections of their own outside Atlassian. Traffic between your browser and your Atlassian site, and between the app's Forge functions and Jira or Confluence, runs over HTTPS (TLS) on Atlassian's platform.
- At rest: Forge storage and its backups are encrypted by Atlassian.
Our own accounts and computers
Aldorock AB is a one-person company. Every company account, GitHub included, requires multi-factor authentication, and passwords are kept in a password manager. The development computer runs macOS with FileVault full disk encryption, the built-in XProtect and Gatekeeper protection, and automatic updates. The apps' code holds no secrets.
Releases and vulnerabilities
Before each release we run the unit tests, forge lint and a dependency scan with pnpm audit, and review what they find. When an advisory sits in a library that comes in through Atlassian's own Forge packages, we update as soon as Atlassian publishes a fixed version.
We follow Atlassian's security bug fix policy. A reported vulnerability is confirmed and scored with CVSS within two business days, and the fix reaches production within these times:
| Severity | Fixed in production within |
|---|---|
| Critical | 10 days |
| High | 4 weeks |
| Medium | 12 weeks |
| Low | 25 weeks |
For a critical vulnerability we also notify affected customers, using Atlassian's vulnerability notification template.
Security incidents
If customer data is exposed or an app is misused, we follow Atlassian's incident management guidelines:
- Within 24 hours of discovery we raise a P1 ticket with Atlassian.
- We contain it by switching off the affected function or rolling the app back to its last good version.
- We update Atlassian at least every 6 hours until the incident is contained.
- We notify affected customers within 72 hours, and the Swedish data protection authority (IMY) where GDPR requires it.
- Within two weeks we review what happened and share the findings on the Atlassian ticket.
Report a vulnerability
Email security@aldorock.com with the app, the steps to reproduce and the impact you see. We acknowledge within two business days. Reports that reach us through Atlassian follow the same process.
What we do not have
- No independent penetration test of the apps has been done.
- We hold no security certifications such as SOC 2 or ISO 27001, and have not completed a CAIQ Lite questionnaire.
- The apps are not enrolled in Atlassian's Marketplace bug bounty program, and we run no bug bounty of our own.
- We run no servers of our own, so there is no log archive or anomaly alerting beyond Forge's own logs.
Changes and contact
Changes to this page are dated at the top. Security reports: security@aldorock.com. Other questions: support@aldorock.com.
