architect
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were detected. The skill is designed to be read-only and explicitly forbids editing files, creating plans, or performing network operations.
- [PROMPT_INJECTION]: The skill instructions do not contain any patterns aimed at overriding agent behavior or bypassing safety guidelines.
- [REMOTE_CODE_EXECUTION]: There are no instructions or patterns for downloading or executing external scripts or packages.
- [DATA_EXFILTRATION]: The skill does not access sensitive system credentials or perform network exfiltration.
- [PROMPT_INJECTION]: The skill reads repository content which is a surface for indirect prompt injection. Ingestion points: repository files read during ground phase. Boundary markers: not present. Capability inventory: read-only access. Sanitization: none mentioned.
Audit Metadata