signup
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious behavior, obfuscation, or sensitive data access detected. The skill consists of markdown-based optimization principles.\n- [INDIRECT_PROMPT_INJECTION]: The skill facilitates reading local project context from files like
.agents/product-marketing.md. While this exposes the agent to data from the project environment, it is used for legitimate context gathering.\n - Ingestion points: Local context files including
.agents/product-marketing.md(SKILL.md).\n - Boundary markers: Not explicitly defined in the instructions for file reading.\n
- Capability inventory: The skill provides instructions for auditing signup flows and does not utilize file system writes or network tools.\n
- Sanitization: Content is used as-is for analysis context.
Audit Metadata