carbon-footprint-tracking

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection where the agent processes external data inputs.
  • Ingestion points: Untrusted data enters the agent context through supplier names, activity descriptions, and material types within the calculator and tracker methods in SKILL.md.
  • Boundary markers: The instructions do not define explicit delimiters to separate user-provided data from agent instructions.
  • Capability inventory: The skill uses Python for data processing and mentions tools for network-based data collection (e.g., requests, beautifulsoup4).
  • Sanitization: No validation or sanitization of input strings is implemented prior to processing.
  • [EXTERNAL_DOWNLOADS]: The skill references several well-known external sources for environmental data and accounting.
  • It identifies reputable emission factor databases including the US EPA, UK DEFRA, Ecoinvent, and the GHG Protocol.
  • It lists common libraries such as requests and beautifulsoup4 as tools for integrating with external APIs and scraping data, which are standard for climate accounting tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 08:18 PM