boxlang-scheduled-tasks
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEPERSISTENCEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PERSISTENCE]: The skill documents the creation of scheduled workloads that are designed to run at recurring intervals. The
bx:schedulecomponent specifically persists task definitions to a configuration file (config/tasks.json), allowing tasks to survive across application restarts. - [DYNAMIC_EXECUTION]: The skill provides instructions for using
schedulerStart(), which dynamically instantiates and executes BoxLang classes based on string-based class names. It also demonstrates the use of.call()to execute functional closures at runtime. - [INDIRECT_PROMPT_INJECTION]: The skill describes an attack surface where external inputs could influence agent behavior.
- Ingestion points: The
bx:schedulecomponent processes remote URLs for HTTP-driven tasks (SKILL.md). - Boundary markers: No delimiters or safety instructions are provided in the examples to prevent the execution of malicious instructions retrieved from these URLs.
- Capability inventory: The skill facilitates full script execution, file persistence, and network requests via the BoxLang scheduler BIFs and components.
- Sanitization: There is no evidence of input validation or sanitization for the URLs or class names provided to the scheduling functions.
Audit Metadata