configuring-filters
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes record data from external systems through extract operations and Handlebars templates, creating a vulnerability surface for indirect injection.
- Ingestion points: Record fields are ingested via the
extractoperation and Handlebarsrecord.syntax as described inSKILL.md. - Boundary markers: No delimiters or boundary markers are mentioned to differentiate instructions from record data.
- Capability inventory: The skill supports both JavaScript-based script filters and Handlebars template execution, which can perform actions based on the ingested data.
- Sanitization: There is no mention of sanitization or validation of record content before it is processed by the filters or templates.
- [DYNAMIC_EXECUTION]: The skill documents the use of dynamic logic through Handlebars templates and custom JavaScript-based script filters.
- Handlebars operands like
{{timestamp "x"}}and{{add record.a record.b}}are executed at runtime to compute filter values. - Script-based filters allow for the execution of custom JavaScript logic referenced by a
_scriptId, which provides a high degree of programmatic control over record processing.
Audit Metadata