testland/qa-hiring-kit
End-to-end structured hiring for QA / SDET / automation / test-lead / quality-manager roles - one chain from job description through interview question bank, competency-anchored scoring rubric, interviewer calibration guide, and post-interview panel debrief, to the 30-60-90 day onboarding plan. Implements the canonical Levashina 2014 et al. structured-interview methodology with ISTQB-aligned competency vocabulary, so the JD a candidate reads, the rubric the panel scores, and the ramp plan the manager runs all describe the same role. Use when opening a QA requisition, authoring any artifact of the hiring loop (JD, questions, rubric, calibration guide, onboarding plan), or running the panel debrief to a defensible hire / no-hire decision.
75
94%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
Why it was flagged
The required workflow involves ingesting interview transcripts, candidate submissions, or scoring rubrics which may contain outsider-authored text during panel debriefs or calibration.
Where we found it
candidate interview transcripts
content-type · 2 sites
The calibration guide requires 2–5 anonymised candidate transcripts as a mandatory runtime input at calibration.md:32, and the calibration session workflow at calibration.md:106 has the panel score a real transcript — both are candidate-authored documents ingested and acted on by the agent at runtime.
references/calibration.md
32
**Sample transcripts** | 2 - 5 anonymised candidate transcripts (or recordings the panel has scored).
references/calibration.md
106
the same anonymised real transcript independently, reveals scores,
interviewer-submitted rubric scores (panel debrief submissions)
content-type · 3 sites
The panel debrief workflow requires each interviewer's independently-submitted rubric scores as a mandatory runtime input that the agent collects, computes agreement on, and uses to drive the hire/no-hire decision.
SKILL.md
83
filled-in rubric with each interviewer's per-dimension scores,
SKILL.md
85
scores were not submitted independently before the debrief
SKILL.md
91
Collect independent scores. Confirm every panelist has scored all dimensions before any scores are shared.
candidate utterances and behavioural observations (quoted during panel debrief)
content-type · 2 sites
The debrief workflow at SKILL.md:103-104 explicitly requires each panelist to quote verbatim candidate utterances and behavioural observations, which the agent then ingests and acts on to resolve scoring disagreements — this is genuine outsider-authored text entering the agent's reasoning via panelist relay, distinct from direct transcript ingestion.
SKILL.md
103
panelist quotes the specific candidate utterance or action that drove their score
SKILL.md
104
(the rubric's anchor principle: anchors describe what the candidate said or did, not what the interviewer felt). Impressions not traceable to a quoted behavioural observation are set aside.
- Audited
- Security analysis