co2-carbon-footprint

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill utilizes standard Python data science libraries to perform its calculations.\n
  • Evidence: Employs pandas for data structuring and openpyxl for generating Excel reports. These are well-known packages from trusted registries.\n- [SAFE]: The implementation follows construction industry standards for CO2 estimation without malicious behavior.\n
  • Evidence: The Python code in SKILL.md defines legitimate classes for life cycle stages and material categories based on EN 15978.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes BIM model data from external files, establishing a standard ingestion surface.\n
  • Ingestion points: The calculate_from_dataframe method reads data from files provided by the user (e.g., bim_quantities.xlsx).\n
  • Boundary markers: While no explicit data delimiters are used in the code, the instructions.md file mandates input validation before processing.\n
  • Capability inventory: The skill utilizes filesystem permissions to read quantities and write Excel reports.\n
  • Sanitization: Input data is cast to float types during calculation, which provides basic structural validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 04:27 AM