Speclet reviews your Jira tickets like a senior QA engineer — drafting acceptance criteria and flagging the coverage gaps your ticket forgot.
Most AI tools generate criteria straight from the ticket. But the ticket is exactly where the gaps live — the permission case, the empty state, the failure path nobody wrote down.
A tool that only rewrites what's already in the ticket can't catch what the ticket forgot. It gives back what you gave it.
A ticket read in isolation misses how it interacts with the epic and the features linked around it.
The missing permission check, the unhandled failure, the race condition — they slip through review and reach production.
Speclet reads the issue plus its parent epic and linked tickets — so it understands how the feature fits the whole.
It drafts Given/When/Then criteria and a test checklist — then flags the coverage gaps your ticket didn't address.
Review, edit anything, and write it into the issue as a comment your whole team can see. It proposes; you approve.
Plenty of tools generate acceptance criteria. Speclet is built to do the harder, more valuable job — telling you what's missing.
The headline feature. Speclet flags the scenarios your ticket didn't cover — permissions, edge cases, failure paths — before they become bugs.
It reads the parent epic and linked issues, not just one ticket in isolation — so criteria account for how features interact.
Built on Forge with Atlassian-hosted AI. Your issue text is processed inside Atlassian's cloud — never sent to a third-party provider or our servers.
It proposes; you approve. Nothing is written silently. And it's a predictable per-seat price — no metered AI credits to watch drain.
Illustrative — final pricing set at launch. Shown in USD, billed by Atlassian on your existing Jira bill.
We're onboarding a small group of early teams before the Marketplace
launch.
Leave your email and we'll reach out.
No spam. We'll only email you about early access.