loopstructural
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill facilitates legitimate scientific modeling workflows using reputable libraries such as LoopStructural, NumPy, and Pandas. All code examples and scripts demonstrate standard data processing and modeling practices without malicious intent.
- [INDIRECT_PROMPT_INJECTION]: The build_model.py script exhibits a standard data ingestion surface that processes external geological datasets. 1. Ingestion points: The load_data function reads user-provided CSV files using pandas.read_csv(). 2. Boundary markers: None present; the script assumes the input is structured geological data according to the documented schema. 3. Capability inventory: The script's operations are limited to mathematical interpolation and writing .vtk visualization files to the local disk. It contains no network access, arbitrary shell execution, or privilege escalation capabilities. 4. Sanitization: The script performs structural validation by checking for mandatory columns such as X, Y, Z, and feature_name.
Audit Metadata