Vera guides
How to point Vera at an app, get tests generated and running, and wire the results into the way your team already ships.
Start here
What Vera is, how to get a first green test, and where it fits in your week.
- Getting startedLink a repository, mint a token, and get a first green test — including how to bring existing Playwright or Cypress specs in without spending AI credits.
- The SDLC loopWhere Vera fits in day-to-day development: the save, feature, PR, bug, morning and release loops, and exactly what executes at each one.
Building tests
The step vocabulary, the builder and recorder, and bringing an existing suite across.
- Actions referenceEvery action a test step can take — clicks, fills, assertions, extraction, network mocking, native gestures — with its fields and the six selector modes.
- Test builderAssembling a test in the visual builder: the action palette, reordering steps, the selector field, running with live per-step progress, and stopping a run.
- RecordingCapture a real browser session as test steps, and the nine-rung priority ladder Vera walks to pick a selector that survives your next deploy.
- Control flowConditional branches and loops: how the flat marker steps pair up, the six condition types, and why AI-generated tests never contain them.
- VariablesThe {{token}} syntax, where values come from and in what order, capturing values mid-run, and why an unresolved token fails the step outright.
- Business contextTell Vera what your product is supposed to do, so generated tests can assert an outcome instead of asserting that something is visible.
- AI planningPlan a fleet of tests before generating it, derive edge cases as datasets, and prove a passing test is wired to your app rather than asserting nothing.
- Repro testsGenerate a test that fails the way a bug report describes, review it before anything is saved, and let a later green run stand as evidence of the fix.
- ChainsRun a test in one project and then a test in another, threading captured values forward — sequential, fail-fast, and masked wherever a value is stored.
- Actor scenariosRun one business story across projects with a different person at each stage — the Customer submits, the Manager approves, the Customer sees the decision.
- Reusable flowsFactor a repeated login or setup sequence into one named flow, reference it with a single step, and let Vera find the duplication for you.
- Import & exportBring existing Playwright, Cypress, WebdriverIO, TestCafe, Postman, HAR or OpenAPI suites in deterministically, plus Java/Python Selenium via AI conversion.
- Test explorerFind your way around a suite that outgrew one page: the health strip, modules, saved views, bulk actions — and where the classification actually comes from.
Checking your app
Beyond pass/fail: appearance, flakiness, and running the suite on a cadence.
- Visual regressionCatch appearance changes assertions miss: baselines, the five diff statuses, the approve-versus-reject decision, and how to stop diffs being flaky.
- Analytics & flakinessPass rates, trends and the flakiness score — how it is computed from the last ten runs, and why a test that always fails scores zero.
- Scheduled runsRun a suite on a cron cadence, how the database-driven tick loop actually fires, and how to choose an interval worth paying for.
- App radarA scheduled structural crawl of your own app that reports what changed and which tests it puts at risk — deterministic, and it never clicks anything.
- Staged repairsThe loop that fixes broken selectors for you: the green gate, what it may never rewrite, and the one switch deciding whether it applies or only proposes.
- Issues & triageFile a bug against a test, run and step, move it through triage, reproduce it as a failing test, and push it to GitHub, Jira or Linear.
- Load & securityPut a path under load with k6 and scan it with ZAP: starting a run, reading p95 and error rate, threshold verdicts, findings, and the honest limits.
On your machine
Run tests against localhost, and give your coding agent credentialed access to them.
- The local agentRun tests on your own machine against localhost with the vera-agent CLI: install, commands, configuration, environment targeting and troubleshooting.
- Claude Code pluginInstall the Vera plugin for Claude Code, what it is for and what it is not, its commands, and the confirmation rule it always follows.
- MCP serverGive an AI coding agent credentialed access to your tests over MCP: minting a token, pointing a client at the server, the tool set, and the limits.
CI & pipelines
Gate merges and releases: pipelines, impact analysis, and your own runners.
- CI integrationRun Vera from GitHub Actions, GitLab CI, Bitbucket Pipelines or any CI system over the raw API, with JUnit XML reports and consistent gating behaviour.
- vera checkTest impact analysis: how Vera maps a diff to the tests that cover it, what map coverage means, the honest limits, exit codes, and how to gate CI on it.
- The CI-riderDrain the run queue from your own CI runner: what --drain does, the GitHub Action, choosing between macOS and Ubuntu runners, and the cost maths.
- GitHub integrationConnect a GitHub account, import repositories as Vera projects, optionally discover their config with AI, and keep them in sync automatically.
Engines & platforms
Browsers, native mobile, and running one suite across a device matrix.
- Browsers & devicesChoosing a browser engine, viewport and device profile, what emulation does and does not prove, timeouts, and real Chrome mode.
- Test enginesChoosing an execution engine, the capability matrix across the three, how selectors translate, known limitations, and how to set up a Selenium grid.
- iOS native testingRun native iOS tests through the local agent: prerequisites, the one-time WebDriverAgent build, choosing a simulator, and .app versus .ipa artifacts.
- Suites & run matrixGroup tests into suites and run them across a device × browser matrix, so one suite covers every combination you care about in a single run.
Running Vera
Operating your own instance, and driving it over the API.
- AI cost & budgetsSee what generation costs before you spend it, bound what one job can spend, and read the numbers correctly — a failed call reports no usage, which is not zero.
- ChangelogWhat changed in Vera and when, newest first — new capabilities, fixes, and changes to how something already behaved.
- Data handlingWhat leaves your machine when you run tests, what a connected GitHub repo sends, what is redacted before it reaches an AI model, and how long any of it is kept.
- Self-hostingRun Vera on your own infrastructure: prerequisites, production build, the configuration that matters, backups, and the two settings to get right.
- APIThe versioned /api/v1 surface for scripts and CI: authentication, running tests, JUnit output, test impact, importing, and the error conventions.