review-mining
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted text from customer reviews, which represents a potential surface for indirect prompt injection. However, the risk is minimal as the script uses deterministic keyword matching and does not perform automated actions based on the extracted text. \n
- Ingestion points: reviews array in the JSON configuration for resources/review-miner.js.\n
- Boundary markers: None explicitly defined in the prompt instructions for review text.\n
- Capability inventory: Statistical calculation, keyword-based theme matching, and extraction of specific sentence types (switching/asks).\n
- Sanitization: Input text is normalized via standard string methods (lowercase, whitespace removal) but not specifically sanitized for LLM control tokens.
Audit Metadata