data-visualization

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes well-established and trusted Python libraries for data visualization (Matplotlib, Seaborn, Plotly, Dash) and data manipulation (Pandas, Numpy). All code samples and scripts demonstrate standard, non-malicious usage patterns.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes functionality to ingest DataFrames and user-supplied metadata (titles, labels) for plotting, and supports saving these plots to the local file system. This represents a functional surface for processing external data, which is inherent to visualization tools and managed by standard library boundaries.
  • Ingestion points: DataFrames and metadata strings in SKILL.md and scripts/quick_plot.py.
  • Boundary markers: None.
  • Capability inventory: Local file writing via Matplotlib and Plotly export methods.
  • Sanitization: Standard library input processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 08:22 AM
Security Audit — agent-trust-hub — data-visualization