
An AI coworker in your tracker, repo, and CI — it writes the test cases from the ticket, triages the flaky failures, and drafts the bug report with the repro steps already in it.
Sauna is an AI assistant for QA engineers that turns the noise into a plan. It writes test plans from the ticket's acceptance criteria, triages CI failures into real regressions versus known flakes, drafts the bug report with repro steps already in it, and watches the flake rate — so your attention goes to the cases that actually break.
What it does
Read the feature ticket and its acceptance criteria and draft a test plan — happy path, edge cases, the negative cases people forget — as a checklist you can run against the build.
Group the run's failures by likely cause, separate the real regressions from the known flakes, and post the summary to the channel before standup.
Turn a reproduction into a structured bug report — steps, expected vs actual, environment, the linked logs — filed to the tracker only when you click create.
A recurring read of which tests are flaking most and which builds went red, so the suite's health is a number you watch, not a surprise.
Put Sauna to work on this.
Get started for freeIn context
Sauna shows up where you already work — the web app, Slack, email, iMessage, and Superhuman. It reads what it needs, does the task, and comes back with the draft for your approval.
Try it
The literal prompt for this job. Open it in Sauna and it picks up from there.
“Read this feature ticket and its acceptance criteria, then draft a test plan covering the happy path, edge cases, and negative cases as a runnable checklist.”
Plugs into the tools you already run — and thousands more, or any MCP server.
What you get
Reports, drafts, and dashboards Sauna assembles from your live tools — kept current on its own.
Get the result, not the busywork.
Get started for freeGood to know
Sauna reads your CI results, triages and explains the failures, and drafts the bug reports and test plans; running the suite stays in your pipeline, and filing waits for your approval.
It groups failures by likely cause and separates the new ones from tests with a known flake history, so you start from a ranked list, not a wall of red.
Sauna reads only what you connect, and acts only after you approve. Your workspace and its memory are yours, not training data.
Keep exploring