security-best-practices
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill defines an indirect prompt injection surface by allowing project-specific documentation to override security rules. \n
- Ingestion points: Project documentation and prompt files located in the user's repository (SKILL.md). \n
- Boundary markers: Absent; the skill explicitly instructs the agent to 'pay attention' to and follow override instructions in external project files. \n
- Capability inventory: The agent has the capability to write security reports to the file system and modify source code to implement security fixes (SKILL.md). \n
- Sanitization: No specific validation or filtering of instructions found in project files is implemented; however, the agent is instructed to report overrides to the user.
Audit Metadata