Everything included. Nothing to maintain.
No scripts to write. No selectors to keep up with. No test code to update when your UI changes. Here's what that gets you.
01
Ship without babysitting a test suite
The maintenance tax on end-to-end tests is the reason coverage slips. BlackBat AI doesn't have one.
No scripts, ever
No test code to author, no selectors to wire up, no config file. You point BlackBat AI at a URL and it works out how to use your app.
Self-adapting to UI change
BlackBat AI tracks behavior, not implementation. A renamed class or restructured DOM doesn't break your coverage.
Nothing to fix after a redesign
No maintenance sprints reconciling a stale suite with a new UI. BlackBat AI re-explores and re-learns the app on every run.
02
Catch the regressions you didn't think to check
Scripted tests only cover paths someone remembered to script. BlackBat AI covers the app.
Autonomous exploration
BlackBat AI navigates your UI like a real user — clicking, filling forms, following links — and maps every page, flow, and reachable state on every run.
Diff runs
Re-crawl only the pages that changed since your last run — a routine re-check after a merge costs a fraction of a full crawl in time and credits, with the same navigation map and findings.
Navigation map
The full behavioral graph of your app in one view — every path, and the gaps in your coverage, before your users find them.
Site documentation and user stories
After a full crawl, BlackBat AI writes up how your app works: what each page does, a description of each section, user stories with success and failure scenarios, and a site overview. Uses credits, so you choose when to run it. Included on plans that list site documentation.
Test coverage for every user story
See which story scenarios have a test plan, link an existing plan, or have BlackBat AI write plans for the gaps, one at a time or in bulk. Plan generation uses credits.
Bugs found in context
Broken interactions, dead clicks, failed flows — each flagged with a screenshot and full reproduction steps, not a red X in a log.
03
Prove the critical flows still work
When you know exactly what must not break, pin it down — in a sentence, not a script.
Plain-English test plans
Describe a journey the way you'd explain it to a teammate — "sign up, verify email, complete onboarding" — and BlackBat AI turns it into a structured, repeatable suite.
Run on demand or on a schedule
Run a suite any time — after a risky merge, before a release, or on a recurring cron schedule. No environment to spin up.
Pass/fail with evidence
Per-step results, and a screenshot plus context for every failure — so a failure points at a specific step, not a vague "the flow is broken".
04
Fit it into how your team works
Shared projects, findings everyone can see, and a path into your pipeline.
Shared projects and findings
One project per app, environments for staging and production, and access for the whole team — so a bug one person triages is visible to whoever fixes it.
Run notifications
Get told when a run completes or issues are detected. Start a run and get back to work.
Test private environments
The BlackBat AI Runner executes from inside your network, so staging and behind-auth apps get tested without being exposed.
Secure by default
Credentials you provide for a run are used only for that run and aren't retained. Your application data stays private.
CI/CD integration
Plug BlackBat AI into your release pipeline, or set up a recurring schedule for crawls and test executions.
MCP integration
SoonConnect BlackBat AI to your AI coding agent via the Model Context Protocol.
BlackBat AI vs. traditional test automation
Both drive a real browser. The difference is what you have to build and keep alive.
BlackBat AI
- Point it at a URL — no setup
- Re-explores and re-learns the app each run
- Coverage grows every run
- UI change is adapted to, not repaired
- Findings arrive with screenshots and repro steps
Scripted frameworks
- Build a framework: drivers, waits, page objects, CI
- Every run replays a fixed, hand-written script
- Coverage grows only where someone writes a test
- UI change breaks selectors and flows
- A failure is a stack trace you go investigate
What's included
Live today, plus what's coming next.
- Autonomous exploration from a URL
- Self-adapting, behavior-based coverage
- Plain-English test plans
- Navigation map
- Diff runs — re-crawl only what changed
- Bug reports with screenshots + reproduction steps
- Visual regression flagging
- Authentication: email/password, magic link, Google
- BlackBat AI Runner for private & staging environments
- Organizations, projects, and team access
- Run notifications
- Scheduled runs — recurring crawls and test executions on a cron
- CI/CD integration via API keys
- SoonMCP integration
- SoonProduct intelligence — spec cross-checking
- Site documentation, user stories and test coverage (plans that include it)