EthicGuard grades every story’s acceptance criteria with AI the moment they’re saved, and keeps the ones nobody can test out of your sprint. Your backlog is read for one call and never stored.
Jira Cloud
LiveVerdicts land on the issue. The sprint gate holds back stories that fail your Definition of Ready.
Start free on Atlassian MarketplaceFree up to 10 users · 30-day trial
Azure DevOps
Early accessThe same reviewer and sprint gate, being built for Azure Boards. Early teams shape what ships first.
Join early accessOne email from the founder when it ships
Pick your role to see what EthicGuard does for you.
Works in Jira Cloud. Azure DevOps is coming on the same grading engine; early access on request.
Works in Jira Cloud. Azure DevOps is coming on the same grading engine; early access on request.
Works in Jira Cloud. Azure DevOps is coming on the same grading engine; early access on request.
Start every story knowing exactly what "done" means.
EthicGuard reads the acceptance criteria before you do, and sends the vague ones back.
of development time is lost to rework. Measured, not asserted.
Besker et al., Chalmers & Oslo, 2019of your backlog content ever stored.
caught defects cover a year of EthicGuard, license and review time included, on a 20-dev team.
Set it up once, then keep writing stories the way you already do.
One click, and it’s free for teams of 10 or fewer.
Your user-story types, your defect types, and the AI key that runs the review.
In the EthicGuard Acceptance Criteria field. That is the whole habit change.
The review runs as you save. A label lands on the issue and a comment names the exact gap.
Other apps generate acceptance criteria. EthicGuard grades them, and keeps unverified work out of your sprints.
A requirements-coverage donut and a weekly defect trend, auto-mounted on a per-project dashboard in Jira.
Unverified stories can’t slip into a sprint. EthicGuard moves them back to the backlog at sprint entry and explains why, and every move is logged.
Every in-scope issue carries a clear verdict label: verified, defect or not ready, stamped right on the issue.
Ambiguity, missing edge cases and untestable assertions, flagged in refinement before anyone writes code.
Catches contradictions between stories before two engineers collide.
Runs on your own Claude, OpenAI or Gemini key, or your own endpoint. Same rubric, same verdicts, zero retention.
Each wedge is one epic, each dot an issue. When two teams write contradictory stories, the link between them turns red before anyone writes code.
Use the arrow keys to pan, Shift and the arrow keys to orbit, plus and minus to zoom, and Escape to reset the view.
Loading interactive map…
Drag to orbit · Ctrl + scroll to zoom · click to focus Shift + arrows to orbit · + / − to zoom · Esc to reset
Developers For a team of devs Loaded cost, per dev per year ($) at loaded each, Stories per dev per month shipping stories per dev a month. EthicGuard bills $9.05 per user a month at that size, free up to 10 users.
EthicGuard ON: caught before code existsFlip it off to compare →
With EthicGuard
$37,333
avoidable rework left, per year
EthicGuard stops about one in three.
Without EthicGuard
$56,000
avoidable rework a year. 120 defects reach your developers.
You get back
$18,667
a year, from 40 defects caught at refinement instead.
Rework left
$37,333
a year, from the 80 defects that still get through.
relative to catching it at refinement
Staff-hours per defect from Wagner, TU München. Field defects run to about 26×; we stop at 12×, and Menzies et al. found the real curve flatter still.
At your numbers, EthicGuard catches about 3.3 requirement defects at refinement each month: $1,555.58 of rework canceled. Twelve of those make $18,667, the You get back figure above.
960 stories/yr ÷ 8 = 120 defects ÷ 3 = 40 caught, 80 still reach development 40 caught × 7 h × $120,000 ÷ 1,800 h = $18,667/yr canceled license $2,172 + 80 h reading reviews ($5,333) = $7,505 all-in → $11,162 net, 149% return 17 catches cover the year
Utrecht University analyzed 1,023 user stories from 18 organizations and found 56% carried at least one defect; their earlier manual pass put it at 24%. We model 1 in 8, well under both, because those studies count structural writing smells, most of which are cheap to fix, and we count only defects severe enough to cause rework. Ours · reasoned, not measured No study reports 12.5%, and we attribute it to nobody. The quantity our model actually needs has never been measured; a rate from real AC-defect labels will replace it and will beat every citation on this page.
Porter, Votta & Basili (University of Maryland, IEEE TSE, 1995) measured requirements inspection finding between 24% and 57% of faults. We sit near the bottom of that band: EthicGuard is an automated grader, not a trained inspection team. Duisburg-Essen ran an LLM over 1,266 annotated Mercedes-Benz requirements and got recall above 0.98 at precision near 0.55: it sees nearly everything and over-reports badly, which is exactly why a human keeps the final say. The other two thirds are the Rework left figure.
Only the difference is counted: writing the criteria costs something either way. A single-digit multiple, not the "100× after release" curve: that curve is challenged by Menzies et al. (NC State / SEI Carnegie Mellon / Fraunhofer, 2017), and the 1:10:100 figure most calculators quote traces to no primary source at all. Wagner (TU München, 2006) puts removal at 1.06 h at requirements review and 8.37 h at system test, a 7.31 h delta, which is where our 7 comes from. Two caveats we would rather state than have found: that paper is a literature survey, not a primary measurement, and it compares techniques rather than following one defect through time. Frattini et al. (2023) reviewed 57 papers and found this cost is reported in 15.8% of them and “never determined empirically”. Nobody has measured it, us included. Hours per developer-year are taken as 1,800, after leave and non-project time.
List price on the Atlassian Marketplace. The ladder is marginal, not flat: a 250-user site pays 100 × $9.05 plus 150 × $7.65, not 250 × $7.65. This page walks that ladder rather than multiplying by a headline rate, so it can never quote a price the listing does not charge, and sites of 10 users or fewer pay nothing at all.
Counted as a cost, because it is one. Against the license alone, the return would read 759%: arithmetically true, and the same meaningless number every vendor calculator produces by dividing savings by list price. Adopting a tool that reviews 960 stories a year is not free because the software is cheap. Five minutes is generous to the cost side: most verdicts are a clean pass. No study publishes this: it is ours to measure from real usage, and it should be replaced the moment we can.
To review a story, EthicGuard sends its text to our analysis engine for a single call, then discards it. We never store titles, descriptions, acceptance criteria or comments. Azure DevOps will ship under the same rule.
Only structured findings (anchors, scores, and message keys) are kept, so the UI re-fetches your live Jira content whenever it renders. If you uninstall the app, there's nothing left to recover. This website uses Google Analytics for aggregate visit statistics; your Jira content is never part of that. Read the full privacy policy.
On Jira, billed through Atlassian alongside your subscription. No separate vendor contract. Azure DevOps pricing comes with its launch.
$9.05 per user / month
Free for 6 months
One requirement defect caught before development typically pays for months of the license.
Three setup steps. Catch your first requirement defect today.
Zero data retention: your backlog content is never stored. How we handle data
“The most expensive bugs I’ve seen weren’t in the code. They were in the ticket, and nobody caught them until a tester asked what done meant.”
Rostyslav Chabria
Founder, EthicGuard · Barcelona
Connect on LinkedIn ↗