voorspellen-dataprep

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions focus on standard data science procedures such as handling missing values, encoding categorical variables, and performing train/test splits using scikit-learn. No dangerous system calls or external network operations were detected.
  • [PROMPT_INJECTION]: The skill processes untrusted user-provided data files, which is an inherent surface for indirect prompt injection. However, the logic directs the agent to use structured programming (sklearn Pipelines) rather than interpreting the data content as instructions.
  • Ingestion points: Step 1 involves reading external files provided by the user.
  • Boundary markers: None explicitly defined in the prompt instructions to delineate data from instructions.
  • Capability inventory: File reading and data manipulation using Python libraries; no capabilities for network access, file writing to system paths, or credential access.
  • Sanitization: The workflow includes specific steps to remove ID-like columns and constant columns, which inherently sanitizes some of the data surface before it is summarized for the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 09:16 AM