director
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and follow instructions from external project files (
AGENTS.mdandCLAUDE.md). This creates a potential surface for indirect prompt injection if these files contain malicious instructions.\n - Ingestion points: Reads policy files at user, project, and working-directory scopes.\n
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the policy files are provided.\n
- Capability inventory: The skill delegates high-privilege tasks including file changes and command execution to other agents.\n
- Sanitization: No specific sanitization or validation of the content within the policy files is mentioned.
Audit Metadata