product-review-extractor
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes product reviews from external e-commerce and review sites, which is a common vector for indirect prompt injection attacks where data can contain instructions that attempt to manipulate the agent.\n
- Ingestion points: Data is ingested from external platforms via tools such as
amazon_bulk_reviews_multi_regionandgoogle_maps_bulk_reviewsmentioned inSKILL.mdandreferences/sandbase-api-map.md.\n - Boundary markers: The instructions do not specify the use of delimiters or clear boundaries to separate untrusted external content from the agent's core instructions.\n
- Capability inventory: The skill relies on
sandbase_call_toolto execute data extraction and aggregation tasks through the sandbase gateway.\n - Sanitization: There is no mention of sanitizing, filtering, or validating the content retrieved from external sources before it is processed by the agent.\n- [NO_CODE]: The skill consists exclusively of instructional Markdown files and configuration metadata, with no accompanying scripts or executable code.
Audit Metadata