kotlin-spring-boot
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The instructions include directives to ignore the model's internal training data regarding software versions, specifically mandating the use of version 4.0.x and 2.3.x regardless of model knowledge (e.g., 'do not downgrade... even if your training data suggests older releases').
- [INDIRECT_PROMPT_INJECTION]: The skill establishes an API surface for processing user-supplied data. 1. Ingestion points: CreateEnvironmentRequest in EnvironmentController.create (SKILL.md). 2. Boundary markers: Uses @Valid and JSR-303 annotations (@NotBlank, @Size) in SKILL.md. 3. Capability inventory: Database operations (EnvironmentRepository) and external client calls (ComputeClient) in SKILL.md. 4. Sanitization: Input validation is provided via annotations, but no explicit sanitization for downstream injection is shown.
Audit Metadata