flink-job-creator
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONNO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill file consists solely of documentation and metadata; no executable scripts, binaries, or configuration files are provided.
- [COMMAND_EXECUTION] (LOW): The skill requests high-privilege tools including Bash, Write, and Edit. Although no malicious commands are defined in the skill itself, these tools represent a significant capability surface if misused.
- [Indirect Prompt Injection] (LOW): Surface for potential data-driven attacks. 1. Ingestion points: User requests for Flink job patterns and code generation. 2. Boundary markers: Absent; no instructions provided to ignore instructions within processed data. 3. Capability inventory: Bash, Write, Edit, Read, Grep. 4. Sanitization: Absent; no verification or escaping logic defined for generated outputs.
Audit Metadata