agency-ai-engineer
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The analyzed skill follows security best practices and contains no malicious instructions or patterns.
- [COMMAND_EXECUTION]: The skill instructions include standard development commands (
cat,ls,grep) used for analyzing local project requirements and metadata. These commands are non-destructive and operate within expected project boundaries. - [PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety filters or override agent behavior. On the contrary, the skill explicitly mandates compliance with AI safety, ethics, and bias-testing standards.
- [DATA_EXFILTRATION]: No attempts to access sensitive system files (e.g., credentials, SSH keys) or perform unauthorized network exfiltration. Mentions of external AI services and frameworks are appropriate for the defined role.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests local markdown files (
requirements.md,data-sources.md), the available capabilities are limited to basic shell utilities, presenting a minimal and safe attack surface.
Audit Metadata