Test cases per criterion
Unit, integration, end-to-end, manual and UAT tests, with Given/When/Then steps and an automation reference, each verifying named criteria.
When an agent writes most of the code, "is this requirement actually met?" needs an answer that does not depend on anyone's memory. Corpole keeps the chain from requirement to acceptance criterion to test to story to pull request, and shows where it breaks.
A criterion is not met because someone ticked it. It is met because its tests pass — and it stops being met the moment one fails.
Unit, integration, end-to-end, manual and UAT tests, with Given/When/Then steps and an automation reference, each verifying named criteria.
CI posts results to one endpoint; the agent reports its own. People record manual and UAT runs with one click.
One failing active test makes a criterion Not met; all passing makes it Met, with the tests as evidence.
A story or task whose covered criteria have a failing test cannot be moved to Done. The message lists the tests and criteria.
Connect the GitHub repository and pull requests and commits that mention a code are linked. Opening a PR moves the story to review, merging it to verification.
The traceability view shows one row per criterion and flags no criteria, no test, no work, no code and failing — with coverage percentages.
A defect is a requirement that is not being met. A risk is a requirement that might not be. Both belong next to the requirement, not in another tool.
Reported by a person or an agent, triaged by severity, tied to the requirement or criterion they break, fixed like a story and proven by a test.
Probability × impact on a 5 × 5 matrix, with a planned response — avoid, mitigate, transfer or accept — and the work that carries it out.
An append-only log of every change to records, criteria, relations, work, tests and code links, filterable by period, type and actor.
Start with one project. The traceability view tells you on day one where criteria have no tests and where tests have no code.