Recap is a Jira Cloud app built on Atlassian Forge. This policy explains what data Recap accesses and, just as importantly, what it does not do with it.
What Recap accesses
When you click “Generate my report,” Recap reads Jira issues assigned to you that entered
a “Done” status in the last 30 days, using the single permission scope
read:jira-work. It reads only the fields needed to build your report: issue
key, summary, parent (epic) key and summary, project key and name, and the
resolution/status-change date. Recap acts as you, the requesting user — it can never see
or report on issues you could not already see yourself.
What Recap stores
Nothing. Recap has no database and uses none of Forge’s storage services. All processing happens in memory for the duration of a single click, and the generated report is returned directly to your browser. Recap keeps no record of your reports, your issues, or your account after the request completes.
Where your data goes
Nowhere outside Atlassian. Recap runs entirely on Atlassian Forge. Report text is generated using Atlassian’s own Forge LLMs API (Atlassian-hosted AI models) — prompts are processed within Atlassian’s platform. Recap makes no calls to any third-party service, uses no external analytics, and has no servers of its own.
Data retention and deletion
Since Recap stores nothing, there is nothing to retain or delete. Uninstalling Recap leaves no residual data behind, because none was ever kept.
Contact
Questions about this policy or Recap’s data handling can be sent to support@isogunlabs.com.