metamorphic-property-extractor

Warn

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local Python scripts (scripts/property_extractor.py and scripts/verify_properties.py) via the command line to process files and generate property mappings.
  • [REMOTE_CODE_EXECUTION]: The core logic of the skill involves performing 'sample executions' of the target program (function.py). If this target program is provided by an untrusted source, the agent may inadvertently execute malicious code during the analysis phase.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the analysis of external code files.
  • Ingestion points: The --program argument accepts user-provided code files (e.g., function.py) for analysis.
  • Boundary markers: No specific boundary markers or 'ignore' instructions are visible in the provided workflow to prevent the model from obeying instructions hidden in code comments.
  • Capability inventory: The skill possesses file system access, local script execution capabilities, and the ability to execute the code it is analyzing.
  • Sanitization: There is no evidence of sanitization or sandboxing for the 'sample executions' mentioned in the workflow.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 6, 2026, 10:20 PM