quarkus-module-build
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to construct and execute shell commands (
find,mvn,cd) using input directly derived from$ARGUMENTS. While the implementation includes a parsing step to split input by spaces, it lacks explicit escaping or sanitization for shell metacharacters such as semicolons (;), pipes (|), or backticks (`). This could allow a malicious module name to execute unintended commands in the local environment. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing untrusted user input through the
$ARGUMENTSvariable and interpolating it into both local shell commands and subagent prompts. - Ingestion points: User input enters via the
$ARGUMENTSvariable inSKILL.md. - Boundary markers: No explicit boundary markers or 'ignore embedded instructions' warnings are used when passing module names to subagents.
- Capability inventory: The skill has the capability to execute shell commands (
find,mvn) and dispatch subagents with specific instructions. - Sanitization: The skill performs basic whitespace splitting and filtering of the word 'and', but does not provide logic to sanitize against shell control characters or prompt-level injection attacks.
Audit Metadata