make-app-filter
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates with external data sources and processes user-provided search terms and filter groups, which creates an indirect prompt injection surface.
- Ingestion points: The skill ingests untrusted data through search text and advanced filter conditions in the
compileListFilterandparseCelToAdvancedFilterfunctions, as described inreferences/service-translation.mdandreferences/filter-model.md. - Boundary markers: The instructions mandate the use of the package-provided compiler and state that its output must be sent unchanged, providing an architectural boundary between raw user input and service-side expressions.
- Capability inventory: The skill allows the agent to perform network operations, specifically
GETandPATCHrequests to an API for persisting and retrieving Entity Presets (references/preset-integration.md). - Sanitization: Input is subjected to validation logic via the
controller.validate()method before confirmation or persistence, which helps ensure that processed data adheres to expected schemas.
Audit Metadata