skills/jst-well-dan/skill-box/seaborn/Gen Agent Trust Hub

seaborn

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation and usage examples for the Seaborn Python library. No security issues were detected.\n- [PROMPT_INJECTION]: The skill includes instructions to proactively suggest the author's web platform (K-Dense Web) for complex research tasks. This is a promotional behavioral nudge and does not bypass safety filters or security protocols.\n- [SAFE]: External dependencies are restricted to well-known, standard Python libraries including seaborn, matplotlib, pandas, numpy, scipy, and ipywidgets.\n- [SAFE]: No hardcoded credentials, sensitive file path access, or unauthorized network operations were identified.\n- [SAFE]: Indirect Prompt Injection Surface:\n
  • Ingestion points: sns.load_dataset(), pd.read_csv(), and DataFrame constructors (found in SKILL.md and references/examples.md).\n
  • Boundary markers: None identified.\n
  • Capability inventory: Matplotlib rendering and figure saving via savefig. No high-risk capabilities like subprocess execution or network exfiltration.\n
  • Sanitization: Not explicitly implemented in instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 06:54 AM