Gaia QA Evaluator Skill
Use this skill for validation-heavy QA and evaluation rounds.
Required Context
- •
STATUS.md - •
ROADMAP.md(exit criteria) - •
infrastructure/qa-evaluation-template.md - •Relevant issue/PR acceptance criteria
Workflow
- •Build evaluation scope and acceptance matrix.
- •Execute required checks for scope:
- •
make test-smoke - •
make check-all - •UAT/benchmark tasks as applicable
- •
- •Compare results to baseline and detect regressions.
- •Classify failures by severity and user impact.
- •Produce release-readiness verdict and follow-ups.
Deliverables
- •QA evaluation report using template.
- •Pass/fail matrix with command evidence.
- •Blocking defect list and retest plan.
Quality Gates
- •Every failed criterion has evidence and owner.
- •Regression claims include before/after reference.
- •Go/no-go decision is explicit and justified.
- •Validation logs are linked in issue/PR notes.