Test executions
Record every test run with a result, build label, executor, duration, evidence and notes — the running history that drives verification status.
Test executions are the running record of verification activity: one row per run of a test case against a specific build. The full history is kept, so you can speak to trends, regressions and "what was the last result on rc-3 versus rc-4".
Where: sidebar Verify → Executions ·
/p/<project>/verification/executions
What you see
- Summary tiles — Total, Pass %, Fail %, Blocked and Skipped, all reflecting the current filters.
- Pass rate by build — a stacked bar chart of passed, failed, blocked and skipped runs per build label, with a pass-rate overlay.
- Filter bar — search across case code, title, executor and notes, plus filters for test case, build label, result, executor and date range. Reset clears them; Export CSV downloads the filtered log.
- Run log — test case, result, build, executed at, executor, evidence count and a notes preview. Select a run to open its detail sheet.
Each run has its own code (for example EXE-000118) and opens at /p/<project>/verification/executions?item=EXE-000118.
Record a run
- Click Record Run.
- Pick the Test case and the Result: passed, failed, blocked or skipped.
- Enter a Build label — a free-form, queryable string such as a release tag (
v1.2.0-rc3), a nightly (nightly-2026-05-22) or a sprint (sprint-12). The build chart groups by exact match, so keep the scheme consistent. - Optionally change the Executor (it defaults to you; pick another member if you ran the test on their behalf), add a duration, paste evidence links (logs, screenshots, reports) and write notes: observed versus expected, anomalies, equipment serial numbers.
- Save.
Recording runs needs edit rights on the project.
How runs roll up
Saving a run updates the requirement's status in the RVTM immediately. Only the latest run of each linked test case counts:
- A failure trumps a pass — if any linked case's latest run failed, the requirement reads Failed.
- A requirement reads Verified only once every linked case has a run and every latest run passed.
The history also rolls up two ways:
- Per test case — passed, failed, blocked and skipped counts and the latest result.
- Project-wide — the same counts across every case, with an overall pass rate.
These summaries feed Verification coverage and Verification reports.
Quick log
For one-off runs against a single case, open it on the Test cases page and use the quick-log buttons. They reuse the last build label the project used.
Filtering recipes
- Compare two builds — read the build chart, or filter by each build and export CSV.
- Find what's red on the latest build — set the build filter to the newest tag and the result to failed.
- Per-executor accountability — filter by executor to see what one engineer ran, useful where independent verification matters.
Co-engineer prompts
Compare build v1.2.0-rc3 with v1.2.0-rc4 and list every test case whose result changed.
Record a passed run of TC-000003 on build nightly-2026-10-27 with the log link I paste next.