About CheckGen

Built for QA engineers who test seriously.

CheckGen started as an internal tool — a way to stop writing the same test cases from scratch every sprint. Paste acceptance criteria, get a grouped checklist, mark outcomes, export. That's it.

Built on Claude Sonnet 4.6 for generation quality, Supabase for real-time shared sessions, and Netlify for hosting. No bloat. Just a focused, thorough testing tool.

Stack
AI modelClaude Sonnet 4.6
DatabaseSupabase
HostingNetlify
Domaincheckgen.dev
By the numbers
2–4×
items per area selected
40
max items generated
6
char share code
$0
to get started
How it works

Three steps to a complete test plan.

01
Paste your AC

Paste your ticket, user story, or acceptance criteria. Full AC works best — a short description still generates useful steps.

02
Pick your areas

Choose which testing areas matter — functional, permissions, edge cases, UI, integrations, error handling, and more.

03
Test and export

Mark items pass, fail, or blocked as you go. Add notes, regenerate any section, and export to CSV when done.

What's included

Everything a test session needs.

🤖
AI from your AC

Grouped test cases in seconds — item count scales with your area selections. Edge cases, validations, break-it scenarios — all from your ticket.

👥
Real-Time Sessions

Share a 6-character code and your team works from the same checklist. Pass, Fail, Blocked sync live.

✎
Notes Per Item

Capture context as you test. Notes export with your CSV so nothing gets lost between sessions.

↺
Regen Any Section

Not happy with a section? Regenerate just that group. The rest of your checklist stays untouched.

📊
Live Progress

Pass, Fail, and Blocked counts update in real time. Time estimates tell you how long a session will take.

↓
Export to CSV

Filter by outcome and export. Notes, priorities, and types all export with each row.

☁
Cloud History

Every checklist auto-saved to your account. Reopen any previous session and pick up exactly where you left off — across any device.

🏢
Team Workspaces

Invite teammates by email, assign roles, and see your team's shared session history in one place. Admin, Moderator, and User roles included.

Features in depth

How it actually works.

AI Checklist Generation

Paste your AC and CheckGen sends it to Claude Sonnet 4.6 with a system prompt that strictly enforces your area selections and scales item count to match — roughly 2–4 items per area selected, up to 40 total. Output is structured JSON grouped by section and rendered into a full checklist.

Real-Time Shared Sessions

When you start a shared session, CheckGen creates a row in Supabase with a 6-character code. Teammates who join poll every 5 seconds and see changes reflected live. No WebSockets required.

History & Cloud Sync

Free accounts store the last 25 checklists tied to your user ID. Each session records full item state, outcomes, notes, ticket ID, and timestamps. Load any previous session and pick up exactly where you left off.

Export to CSV

Filter by outcome and export. Each row includes item text, section, priority, type, time estimate, outcome, and notes. Paste straight into your defect tracker or attach to a ticket.

Pricing

Free to start.

Start with a free account for cloud history, shared sessions, and team workspaces — no credit card required.

No Account
Free

Try it without an account.

  • 3 AI generations
  • Personal sessions only
  • Browser-local history
  • Shared sessions
  • Cloud history
  • CSV export with notes
Try Now →
Recommended
Free Account
$0 / month

Everything you need for day-to-day QA testing.

  • Unlimited AI generations
  • Personal & team sessions
  • Cloud history (last 25 checklists)
  • CSV export with notes
  • Team workspace & invite links
Create Free Account →
FAQ

Common questions.

Does CheckGen store my ticket data?
Ticket text and AC you paste is sent to the Anthropic API to generate your checklist. It is not stored on CheckGen servers unless you save a session — in which case the checklist items and metadata are stored in your Supabase-backed account, not the original ticket text.
What AI model does it use?
Claude Sonnet 4.6 via the Anthropic API. Sonnet produces more specific, executable test cases than lighter models — generation typically takes 15–60 seconds depending on ticket complexity and the number of testing areas selected.
Can I use it without Jira or a ticket system?
Yes. CheckGen works entirely standalone. Ticket ID and environment fields are optional. Just paste your AC and go.
How does shared session sync work?
Shared sessions use Supabase REST API. Every outcome change pushes a PATCH to the session row. Teammates poll every 5 seconds and see changes reflected live. No WebSocket connection required.
How long do shared sessions stay active?
Sessions stay active until someone clicks End Session or the row expires after 30 days of inactivity. There is no time limit during an active test run.
What happens to anonymous history?
Anonymous history is stored in your browser localStorage. It disappears if you clear your browser data. Create a free account to save sessions to the cloud permanently.
Do teammates need an account to join a shared session?
Yes. Shared sessions require a free CheckGen account. This lets the session track who marked each outcome and saves a copy to each participant's history. Creating an account takes under a minute and is completely free.
What's the difference between Personal and Shared sessions?
Personal sessions are just for you — checklist saved to your cloud history, no join code. Shared sessions generate a live session your teammates can join by code or invite link. Outcomes sync in real time, and a name pill shows who marked each item. Shared sessions are saved to both your history and your team history.
How do team workspaces work?
Create a workspace from the Team page and invite members by email. Three roles are available: Admin (full access including settings and roles), Moderator (can view team history but not change settings), and User (can only see their own history). All shared sessions created by workspace members appear in Team History.
Why does item count vary between runs?
Item count scales with the number of testing areas you select — roughly 2–4 items per area. Selecting 3 areas produces a focused 6–12 item checklist; selecting 8 areas produces a more thorough 16–40 item checklist. This means you're always getting appropriately sized output for the scope you asked for.