test-gaps

name	test-gaps
description	Compare changed behavior and risk hypotheses with existing automated checks to find important unverified outcomes and boundaries. Use when planning verification for a change, reviewing test adequacy, or deciding whether a temporary generated probe is justified.

Test Gaps

1. List the changed observable behaviors, failure modes, and boundary conditions from repository evidence.
2. Inventory relevant existing checks from configuration, CI, task files, manifests, scripts, and documentation.
3. Map each behavior to checks that exercise it through a consumer-visible surface. Do not credit tests that only restate implementation details.
4. Record uncovered high-risk behaviors, including invalid input, empty and boundary values, interruption, persistence, Unicode, permissions, and external failures when relevant.
5. Prefer extending or invoking an existing check. Request a temporary probe only when no existing check can observe the hypothesis.
6. Keep probes read-only with respect to the repository and place their source under the allocated artifact directory.

Required evidence

For every gap, provide:

• the uncovered behavior and why it matters;
• existing tests inspected and why they do not cover it;
• the smallest exact command or probe that would cover it;
• the expected observable result and artifact;
• whether missing evidence would make the run inconclusive.

Do not infer coverage from filenames or test counts. Read the behavior asserted by the check.