About
What is Tada?
Tada is an automated QA assistant for AI-written code. It signs in as each of your users, checks the feature in a real browser, and reports Passed or Failed. With GitHub connected, the verdict lands on the pull request as a check. Every check opens as a report. Its home is taada.dev, spelled t-a-a-d-a dot dev, with two a's.
What Tada does
- You paste your app's link, then name a feature and the people who use it. GitHub is optional: connect it to add pull request checks and let Tada read your code.
- Tada writes the tests for that feature and keeps them. Nothing lands in your repository.
- Tada checks the feature when you press Check now, when your coding agent asks, or, with GitHub connected, when a pull request deploys.
- Tada signs in as everyone the feature needs, each in a browser of their own, and goes through it one step at a time.
- Each step names something that must appear on screen. If it never appears, the check is Failed.
- The check opens as a report: every test's result, the step that stopped a failed one, and the film.
- A check that passed gives the feature its demo, which anyone with the link can watch without signing in.
On a paid plan, mark the check required in branch protection and a Failed check stops the merge, the same way your CI already does. On the Free plan the check is posted as a warning. People still decide what ships, and you can always override.
The words Tada uses
- An app has features. A feature has tests, and it is Passing or Failing.
- A check is one go at a feature. It ends Passed or Failed, and it opens as a report.
- The cast is the people Tada plays, such as an admin and the teammate they invite. Each actor has a name, a sign-in and a voice.
- A feature's demo is the film of its newest passing check.
- Tasks are the few things only you can do for Tada, such as a sign-in. Tada lists them and clears them by itself.
When Tada checks
- Each time a pull request deploys to staging or a preview, for an app connected through GitHub.
- When you press Check now on a feature, or ask Rivet, the QA colleague inside Tada.
- When your coding agent asks, including for an app on your own machine.
Tada can also join a meeting and present a demo. Calls are being tried in a few workspaces first.
Who it is for
Founders and small teams shipping code written with AI tools such as Claude Code, Cursor and Copilot, with no QA hire. A team that has a QA gets a second pair of hands.
What Tada is not
- Not a code reviewer. It does not read your diff. It opens the app and uses the feature.
- Not a test runner for your CI alone. The result is made to be shown: a film for a pass, the step where it stopped for a failure.
- No AI decides the result on its own. A check follows each test's steps as written, so the same test gives a comparable result every time. A step marked as a judgment is answered by a model, and the report says so.
- Not a promise that nothing breaks. Tada checks the features you named, and tells you which ones it did.
Where the checks happen
Tada's servers do every check. If your app is on your own computer, the Tada tool opens a private link to it for the length of the check so Tada can reach it.
What it costs
You pay per check, never per person, so your whole team can watch for free. The Free plan includes 40 checks a month. See pricing for the paid plans.
Other things called Tada
Other projects share the name and are not connected to this one: TaaDa, an Android Auto app for Tesla cars, and TAADA, a tool used in reading research.
Who makes it
Tada is made by Modulab, a small software studio. Write to [email protected] and a person reads it.