multi-persona-review
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill does not contain any code, scripts, or command-line instructions, eliminating the risk of remote code execution or unauthorized system access.
- [SAFE]: No network operations, external downloads, or data exfiltration patterns were detected; all operations described are internal to the agent's reasoning process.
- [SAFE]: No obfuscation techniques, hidden URLs, or base64-encoded strings were found in the skill's markdown or configuration files.
- [PROMPT_INJECTION]: The skill ingests untrusted data in the form of 'artifacts' for review as specified in Step 1 of SKILL.md. Boundary markers and sanitization are absent. However, the skill possesses no dangerous capabilities such as file writing, network access, or subprocess execution across any of its files (SKILL.md, agents/openai.yaml, references/reviewer-design.md), rendering the surface for indirect prompt injection safe.
Audit Metadata