real-estate-debt-analysis
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No prompt injection or safety bypass instructions were found in the skill's metadata or command triggers.- [SAFE]: The skill ingests data from multiple surfaces including Excel, CSV, JSON, and extracted text from images and audio. Analysis of scripts/real_estate_analysis.py shows that data is processed using safe parsing libraries (csv, json) and basic type validation without any dynamic execution or unsafe interpolation into system commands.- [SAFE]: Network operations are handled via environment variables (REPORT_GENERATOR_SERVICE, DATA_PROCESSING_SERVICE), which is a standard pattern for configuring external microservices in AI agent skills. No hardcoded or unauthorized external domains were detected.- [SAFE]: The skill's dependencies, requests and pandas, are reputable libraries used for their intended purposes. No unauthorized remote code downloads or execution patterns were identified.- [SAFE]: There is no evidence of hardcoded credentials, sensitive file access (e.g., SSH or cloud config), or persistence mechanisms. The use of a local configuration file for price data is a standard functional requirement.
Audit Metadata