write-script-java

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to execute shell commands using the wmill CLI to manage the lifecycle of Java scripts, including local previews, remote execution, and deployment to a workspace.
  • [EXTERNAL_DOWNLOADS]: Uses the wmill generate-metadata and wmill script preview commands to resolve and download Maven dependencies defined within Java script comments (e.g., //requirements:).
  • [PROMPT_INJECTION]: Contains a potential surface for indirect prompt injection where user-provided inputs are interpolated as arguments for script preview commands (wmill script preview <path> -d '<args>').
  • Ingestion points: User-provided values for script parameters in the -d argument.
  • Boundary markers: Absent; the command uses standard shell quoting which may be insufficient for complex adversarial inputs.
  • Capability inventory: Shell command execution via wmill, file system writes for metadata, and network operations for deployment.
  • Sanitization: The instructions recommend using "plausible args" but do not provide specific sanitization or escaping guidelines for the shell interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 06:29 AM
Security Audit — agent-trust-hub — write-script-java