scientific-visualization
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted external data (e.g., pandas DataFrames) to generate visualizations, which presents a surface for indirect prompt injection.
- Ingestion points: Data enters the context through plotting functions such as sns.scatterplot and ax.plot as shown in SKILL.md and references/matplotlib_examples.md.
- Boundary markers: No explicit boundary markers or instructions are provided to delineate untrusted data from the agent's instructions within the plotting context.
- Capability inventory: The skill has file-writing capabilities through fig.savefig in scripts/figure_export.py and create_style_template in scripts/style_presets.py.
- Sanitization: No sanitization or validation of strings used for titles, labels, or legends is performed in the provided scripts.
Audit Metadata