re-engagement
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data about dormant leads and inactive customers to generate outreach sequences, creating a surface for indirect prompt injection.
- Ingestion points: User-provided lead names, original assets, and customer history used in campaign design.
- Boundary markers: Absent in the output templates, which does not isolate user-supplied strings from the prompt context.
- Capability inventory: Minimal; the skill involves text generation and local validation via
scripts/check-output.pywith no network or sensitive file system access. - Sanitization: No sanitization or validation of the content of user-supplied lead data is performed before interpolation into templates.
Audit Metadata