Use at the start of every QA round - recording the case matrix with record_test_cases (categories, criterion links, rejected cases) before touching the app, then executing all of it
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
Before you boot anything or call anything, derive the cases — from the task description and its acceptance criteria, never from the code or the diff (techniques and sources: test-scenario-design) — and write them onto the task with record_test_cases (status=planned). Never post the plan as a comment; the card is where it lives.
The criteria are the floor, not the ceiling. They summarise the request in a few lines, so start from them and then ask what the person asking would expect to be true once this is built:
criterion_id.criterion_id empty (boundary-negative-testing for the data catalogue).design: cases per state and width (design-conformance).status=invalid with the reason in notes.Then execute the matrix in this same run and record every verdict (set_test_case_result): passed, failed with what you actually observed, or skipped with what blocked it. A case still planned when you try to hand the task on refuses the move.
The category enum has no accessibility/performance/security/contract value of its own — map them: a11y → visual + title prefix a11y:; performance → other + perf:; authz/security → auth; contract/compatibility → negative or happy_path depending on the case; concurrency/duplicate submission → negative; worker/async → async.
<surface>: <action> -> <expected>, e.g. settings: rename to 81 chars -> 422. Titles are the case's identity across rounds — keep them stable so a re-test (retest-after-revision) matches the right history.
Is not a case you invent a proxy for. When there is a concrete choice between two readings, ask_user with the readings as options and leave it unapproved meanwhile. When it is unclear with no concrete reading to offer, reject it via review_criterion ("untestable as written: <why> — needs a sharper criterion") so the move to need_revision puts it in front of the human — a comment asking the PM to sharpen it reaches nobody, since the PM does not subscribe to in_qa.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Use when writing acceptance criteria for a task - express each as an observable Given/When/Then that QA can execute, including negative cases
The English summary is being prepared. Showing the original description.
Use when the diff adds or changes an endpoint, resolver, RPC, job or query that takes an object id, a role check, a request binding or a tenant filter - BOLA/IDOR, function-level authorization, mass assignment and tenant scoping
The English summary is being prepared. Showing the original description.
Use on every UI change - semantic HTML, labels for controls, keyboard-navigable dialogs/menus, visible focus, and never color as the only signal
The English summary is being prepared. Showing the original description.
Use when a task changes any screen, form, dialog, menu or control - Lighthouse/axe scan of the changed screens, a keyboard walk, and the thresholds that fail a task
The English summary is being prepared. Showing the original description.
How to work a task returned with review, QA or UAT findings. Use when a task is in need_revision or PR review comments are in your context.
The English summary is being prepared. Showing the original description.
Use when the repository keeps its alerting in code, or a person asks for alerts outside TaskTrooper - PrometheusRule YAML, a Cloud Monitoring alert policy in Terraform, a Datadog monitor, each routed to TaskTrooper's alert ingest URL so it opens an incident here
The English summary is being prepared. Showing the original description.