branding
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [DATA_EXPOSURE]: The skill instructions include steps to read project-specific files such as
.claude/project-context.mdand.cursor/project-context.md. These files are standard metadata locations for development-oriented AI agents to understand project goals and do not contain system-level credentials or sensitive user secrets.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external documents (such as brand PDFs or lists) provided by the user, which establishes a theoretical attack surface for indirect prompt injection. - Ingestion points: User-provided PDFs or text lists ingested during the Visual Specification Delivery process.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the external content.
- Capability inventory: The skill does not utilize high-risk tools; it lacks commands for network exfiltration, filesystem modification, or subprocess execution.
- Sanitization: No validation or sanitization logic is present for the ingested data.
Audit Metadata