Identify acceptance gaps and generate explicit QA verdicts from available evidence.
npx skills add https://github.com/almazom/agents_slash_skills --skill saad
Please help me install this Agent Skill. Skill: saad Source: https://github.com/almazom/agents_slash_skills/tree/main/saad Command: npx skills add https://github.com/almazom/agents_slash_skills --skill saad
Saad provides a QA-centric decision framework to determine if acceptance criteria are truly met, by isolating evidence, guiding minimal reproduce steps, and delivering explicit verdicts.
Ask Saad to assess the current evidence and return a clear verdict with the strongest proof path.
High-intent search queries and answers about installing and using this skill.
Generate explicit QA verdicts by isolating available evidence and checking it against bounded acceptance criteria. This framework outputs ACCEPTED, REJECTED, or INCONCLUSIVE decisions alongside minimal reproduce steps and proof collection artifacts.
A minimal reproduce flow in QA proof collection is a guided path to verify claims using the strongest available evidence. It helps isolate acceptance gaps and generate explicit verdicts by focusing on reproducible artifacts across pane or parallel tasks.
Identify acceptance gaps in bounded checks by assessing the current evidence against required criteria. This QA framework isolates missing proof, returns an INCONCLUSIVE or REJECTED verdict, and guides the next steps for evidence collection.
Yes, you can use this QA verdict framework for parallel tasks. It supports cross-pane coordination and escalation boundaries to maintain accountability while collecting proof and generating explicit ACCEPTED, REJECTED, or INCONCLUSIVE verdicts.
Use an INCONCLUSIVE verdict in QA testing when available evidence is insufficient to fully meet bounded acceptance criteria. This verdict shape indicates identified acceptance gaps and guides the next steps for minimal reproduce and proof collection.
wsmanReview test files and manual evidence for coverage gaps before QA sign-off.
baphledEvaluate change requests and review comments for validity and architectural alignment.
aayushsoamCoordinate parallel AI agent sessions across multiple harnesses using dmux pane manager.
servitolaRun unit, integration, and E2E tests and verify acceptance criteria to generate a deploy readiness verdict.
