screen-check
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read external project files, which represents a potential surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: The agent reads local project files including
DESIGN.md,AGENTS.MD, and changed source files (views, templates, styles). - Boundary markers: No explicit delimiters or boundary markers are defined for the content being read.
- Capability inventory: The skill is strictly limited to providing textual reports and review findings. It explicitly forbids network operations, software installation, and creating new files.
- Sanitization: No specific sanitization or filtering of the input file content is mentioned.
- [SAFE]: The skill references documentation from Vercel Labs' public repository as a reading pointer. This is a reference to a well-known service and the skill explicitly instructs the agent not to fetch the URL at runtime.
Audit Metadata