feature-gating
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends installing the
skene-skillsutility from the vendor's official GitHub repository (github.com/SkeneTechnologies/skene-skills) to facilitate growth audits. It also references an external blog post on Amplitude for methodology context. - [INDIRECT_PROMPT_INJECTION]: The skill contains a 'Codebase Audit' section that instructs the agent to scan local files for specific string patterns related to subscription tiers and feature flags (e.g.,
plan,tier,paywall). This creates an ingestion surface for untrusted local data. - Ingestion points: User's local codebase files (referenced in SKILL.md under 'Codebase Audit').
- Boundary markers: Absent.
- Capability inventory: File-read access via searching tools. No write or network operations are instructed following the audit.
- Sanitization: Absent; the agent is directed to report found patterns directly.
Audit Metadata