paywall-upgrade-cro

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends the installation of skene-skills via a link to https://github.com/SkeneTechnologies/skene-skills. While this repository belongs to the skill's author, it represents an external dependency that the user is encouraged to incorporate into their environment.
  • [INDIRECT_PROMPT_INJECTION]: The 'Codebase Audit' section instructs the agent to search for and read various files within the user's repository (e.g., component files, trigger logic, and analytics).
  • Ingestion points: Content from files found via searches for patterns like *paywall*, *upgrade*, and *gate* in the user's codebase.
  • Boundary markers: The instructions do not specify any delimiters or warnings to ignore instructions embedded within the codebase being audited.
  • Capability inventory: The agent is prompted to perform searches and inventory touchpoints, which requires file system read access.
  • Sanitization: There are no instructions for sanitizing or filtering the content read from the user's files before processing it.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 02:47 PM