writing-skills
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The
render-graphs.jsscript usesexecSyncto run the system'sdotcommand (Graphviz) for rendering diagrams. This is a local utility function intended for visualization within a development environment. - [EXTERNAL_DOWNLOADS]: The documentation references and recommends the installation of several well-known third-party libraries such as
pdfplumber,docx-js, andpypdffor various file processing tasks. These are standard packages in their respective ecosystems. - [PROMPT_INJECTION]: The skill provides templates for 'pressure scenarios' designed to test agent compliance with documentation. These templates use imperative language and role-play triggers (e.g., 'IMPORTANT: This is a real scenario') which are characteristic of prompt injection techniques; however, they are used here as legitimate tools for testing and debugging agent behavior.
Audit Metadata