performance-review
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists of natural language instructions for conducting performance reviews. No malicious code, obfuscation, or dangerous capabilities were detected.
- [NO_CODE]: No scripts, commands, or executable logic are present in the skill.
- [PROMPT_INJECTION]: The skill instructions involve gathering and processing external peer feedback and performance evidence. This creates a surface for indirect prompt injection where malicious text in the source data could attempt to influence the agent's behavior. However, the risk is negligible as the skill has no tool access or dangerous capabilities. (Evidence Chain
- Ingestion points: Peer input and performance evidence gathering in SKILL.md; Boundary markers: Absent; Capability inventory: None; Sanitization: Absent).
Audit Metadata