academic-figure-drawing-harness
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external scholarly content and user feedback to drive figure generation, which creates an indirect prompt injection surface.
- Ingestion points: Scholarly literature and articles (e.g., PubMed unified_search) and user feedback.
- Boundary markers: None specified to delimit external content within prompts.
- Capability inventory: File modification via 'replace_string_in_file' and Python-based script generation for plotting.
- Sanitization: No sanitization or validation of input content is described.
- [COMMAND_EXECUTION]: The skill generates and executes Python scripts using matplotlib and plotly to render structured data visualizations.
Audit Metadata