refactor
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external source code provided through the
$ARGUMENTSparameter. This creates a surface where malicious instructions hidden within comments or data in the source files could influence the agent's behavior. - Ingestion points: The files, directories, or modules specified in
$ARGUMENTSand identified during the scope phase. - Boundary markers: The instructions lack explicit boundary markers or directives to the agent to disregard natural language instructions found within the source code.
- Capability inventory: The agent has the ability to read and write files, as well as execute local commands (e.g., test runners).
- Sanitization: There is no evidence of sanitization or safety checks performed on the content of the ingested files.
- [COMMAND_EXECUTION]: The skill workflow includes a mandatory step to "Run the suite," which involves executing local shell commands to trigger the project's test runner. This is a functional requirement for refactoring but provides a path for code execution within the environment.
Audit Metadata