subscription-benchmarks
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided performance metrics and monetization queries to select the appropriate benchmarking skill.
- Ingestion points: User benchmark data and monetization questions are processed as input for the routing logic.
- Boundary markers: The skill does not implement delimiters or specific instructions to isolate user data from the agent's control flow.
- Capability inventory: The skill references and coordinates other reporting skills but does not request direct file system or network tool access in its metadata.
- Sanitization: The instructions do not define methods for validating or sanitizing user-supplied metrics before they are used to query downstream benchmarks.
- [EXTERNAL_DOWNLOADS]: The skill documentation includes a link to an external repository on GitHub (https://github.com/vkharlan-adapt/state-of-paywalls-2026-skill) for a related skill component.
Audit Metadata