The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several Node.js scripts intended for slide automation and quality auditing.
Scripts like render_slides.mjs and export_ppt.mjs use the standard node:util and node:fs/promises modules to manage file operations within the project's output directory.
The build system uses child_process.spawn to coordinate the rendering and export pipeline, which is a common and legitimate pattern for development tools.
[EXTERNAL_DOWNLOADS]: The skill utilizes legitimate external resources for rendering and styling.
It requires the installation of Playwright browser binaries via npx playwright install chromium, which is standard for web-to-image conversion.
Style specifications (style_a.md through style_j.md) reference Google Fonts for typography, which are considered safe and well-known external assets.
[PROMPT_INJECTION]: The SKILL.md file and associated reference guides provide detailed task-oriented instructions for the agent.
No override markers, bypass attempts, or malicious role-play patterns were detected.
The instructions explicitly restrict the agent from using JavaScript in the generated slide HTML, serving as a proactive security boundary.
[DATA_EXFILTRATION]: All data processing and file creation occur within the local environment.
No unauthorized network communication or attempts to access sensitive system files (e.g., credentials, SSH keys) were found during the analysis of the scripts and workflows.

ppt-design