fishbone-diagram
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (e.g., problem statements and causes) and renders it into SVG diagrams and HTML reports. \n
- Ingestion points: User-supplied text via chat is processed and saved to fishbone_data.json or passed directly to scripts for visualization. \n
- Boundary markers: SKILL.md contains explicit instructions to treat all user-provided text as data, not instructions, and warns against interpreting cause descriptions as agent directives. \n
- Capability inventory: The skill uses Bash to execute local Python scripts which can read and write files (JSON, SVG, HTML). It does not have network access or dynamic code evaluation capabilities. \n
- Sanitization: The scripts generate_diagram.py and generate_report.py implement escape_xml and escape_html respectively to sanitize user-provided content before rendering. \n- [SAFE]: The skill implements robust security practices for a developer-oriented tool. \n
- The Python scripts include path validation logic (_validate_path) that prevents path traversal attacks and restricts file output to specific extensions (.json, .svg, .html). \n
- All processing occurs locally using standard Python libraries; no external network requests or third-party dependency installations are performed.
Audit Metadata