EpicStory for Jira
Turn Jira Epic and Project requirements into editable AI-draft stories and subtasks with Atlassian-hosted Forge LLM and human review before publishing.
EpicStory runs inside Jira Cloud on Atlassian Forge. It converts Epic descriptions and Project requirements into editable AI drafts using Forge LLM. Claude Haiku 4.5 is the model currently available to EpicStory; the model may change as Atlassian updates Forge LLM availability. Teams review and edit every draft before choosing what to publish.
See EpicStory in Jira
This walkthrough shows an earlier EpicStory release turning an Epic into reviewable Jira stories and publishing only approved work.
- Generate a first backlog draft directly from an epic
- Review and edit stories before anything is published
- Publish approved work items back into Jira with the right structure
Product walkthrough from an earlier release.
Core Capabilities
Epic and Project Generation
Generate structured Jira drafts through an asynchronous pipeline that extracts requirements before producing stories and subtasks.
Draft Workspace
Every run becomes a selectable draft with a tree view for stories, subtasks, priorities, story points, and labels, so teams can edit before a single issue is published.
Full Project Draft Regeneration
Regenerate an active Project draft using the current draft and a new instruction. Issue-panel partial refinement is not available in version 7.
Guided Publishing
A publish panel summarizes what is about to ship, then calls secure resolvers that create issues, link them back to the source epic, set story points, and ensure the correct Jira issue types exist.
Use Cases
- Stand up a release backlog directly from an epic
- Create reviewable first drafts for regulatory or security initiatives while retaining human approval
- Prep delivery teams by exporting final stories, subtasks, and acceptance criteria straight into the Jira epic they will execute
Integrations & Architecture
EpicStory runs inside Jira Cloud on Atlassian Forge. Production AI generation uses Atlassian-hosted Forge LLM, currently with Claude Haiku 4.5. No external AI or observability egress is configured in production. Forge-hosted storage supports Atlassian data residency controls; transient runtime and model routing are subject to Atlassian's platform operation and are not represented as a guaranteed inference region. Data handling details are documented in our legal policies.
Pricing and Trial
EpicStory is paid via Atlassian Marketplace. Marketplace may offer a trial for Jira Cloud, but AI generation is disabled during the trial because Forge LLM calls are paid. An active paid license is required to generate new drafts. The Atlassian listing is the authoritative source for pricing, billing, and renewal terms.
- Marketplace trial supports evaluation without paid AI generation
- Marketplace pricing and billing are the live commercial source of truth
- Support, privacy, and service-availability terms are available before rollout
Need procurement or rollout context before installing? Review our privacy and trust resources or contact us.
Ready to try EpicStory in Jira?
Review the Marketplace listing or contact us if your team needs rollout support.
AI generation is disabled during the Marketplace trial and requires an active paid license.