harden-for-production
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied application configuration and code snippets, creating a surface for indirect prompt injection. * Ingestion points: User-provided NestJS application code and configuration queries described in SKILL.md and evals/skill.json. * Boundary markers: None identified; instructions do not require delimiters for untrusted data. * Capability inventory: Modifications to application entrypoints (main.ts), module configurations, and environment variables. * Sanitization: No specific sanitization or validation logic is defined for inputs.
- [EXTERNAL_DOWNLOADS]: The skill references the installation of @eleven-labs/nest-profiler and @eleven-labs/nest-profiler-ai via standard package managers. These are official vendor-managed Node.js packages.
Audit Metadata