pseudocode-to-java-code

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill is entirely composed of Markdown documentation and a static Java template (assets/java_template.java). It does not contain any executable scripts or automation logic, which eliminates typical attack vectors like remote code execution or data exfiltration.
  • [PROMPT_INJECTION]: The skill processes user-supplied pseudocode, creating a surface for indirect prompt injection. 1. Ingestion points: User input during the workflow in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: None. 4. Sanitization: Absent. Despite the surface, the lack of internal capabilities makes this a safe reference tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 10:21 PM