dwg-to-excel

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of the DwgExporter binary via subprocess.run. This is the intended mechanism for processing DWG files and is used specifically for the conversion task using a list-based command structure, which minimizes shell injection risks.
  • [PROMPT_INJECTION]: The skill processes user-provided DWG files and extracts text content such as annotations and attributes, creating a surface for potential indirect prompt injection.
  • Ingestion points: Content is read from generated Excel files via read_entities and get_text_content methods in the DWGExporter class.
  • Boundary markers: None identified; extracted text content is processed and presented to the model as standard data.
  • Capability inventory: The skill possesses filesystem read/write access and the ability to execute the DwgExporter binary.
  • Sanitization: No explicit sanitization or filtering of the text content extracted from the CAD drawings is implemented before processing.
  • [SAFE]: No evidence of malicious activity was found. The skill does not contain hardcoded credentials, perform suspicious network requests, or use obfuscated code. It strictly adheres to its stated purpose of CAD data extraction and analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 05:43 AM