qa-auditor
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and execute dynamic testing on outputs from other agents, creating an attack surface for indirect prompt injection.
- Ingestion points: Artifacts or systems ready for audit, outputs of all other agents (SKILL.md)
- Boundary markers: None present in instructions
- Capability inventory: Static analysis, linters, and dynamic testing (SKILL.md)
- Sanitization: No sanitization or validation steps mentioned for incoming content.
- [NO_CODE]: The skill consists entirely of instructional markdown and does not contain any executable scripts or binary files.
Audit Metadata