writing-skills
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The provided Node.js utility
render-graphs.jsexecutes thedotbinary from the Graphviz suite and thewhichcommand usingchild_process.execSync. These executions are used to render diagrams extracted from markdown and do not involve untrusted shell interpolation. - [EXTERNAL_DOWNLOADS]: The documentation mentions several reputable Python packages (
pdfplumber,pypdf,pytesseract,pdf2image) and the system toolgraphviz. These references are provided as examples of dependencies for specific tasks and represent well-known, trusted resources. - [PROMPT_INJECTION]: The skill includes extensive documentation on 'Persuasion Principles' and 'Pressure Scenarios' used to influence and verify agent behavior. While these methodologies utilize psychological framing similar to prompt engineering techniques, they are intended for improving the reliability and discipline of agent skills during the authoring and testing process.
Audit Metadata