paywall-upgrade-cro
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains purely instructional content for conversion rate optimization (CRO) in software products. It outlines best practices, component checklists, and experiment ideas without executing any code or performing network operations.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions include a directive to read a local file,
.claude/product-marketing-context.md, to gather product information. This represents a potential surface for indirect prompt injection if the content of that file is controlled by an external party. However, as the skill lacks capabilities for command execution, network exfiltration, or privileged file writing, the risk associated with this ingestion point is negligible.
Audit Metadata