The Agent Skills Directory

[SAFE]: The skill follows a legitimate and well-structured workflow for generating knowledge comics from user-provided content without any detected malicious intent.\n- [COMMAND_EXECUTION]: The skill uses the bun runtime to execute a local script (scripts/merge-to-pdf.ts) for PDF generation and suggests system-level tools like pngquant, optipng, or sips for image optimization.\n- [EXTERNAL_DOWNLOADS]: The skill requires the bun and npx runtimes as specified in its metadata and utilizes the pdf-lib package within its PDF merging script.\n- [PROMPT_INJECTION]: The skill processes external text material, which creates a surface for indirect prompt injection. Ingestion points: user-supplied source files (source.md). Boundary markers: structured storyboard and prompt templates. Capability inventory: local file read/write operations, script execution via bun, and delegation to external image generation skills. Sanitization: implicit filtering occurs through the AI-led analysis and simplification phases. No direct or malicious injection attempts were detected.\n- [DATA_EXFILTRATION]: No evidence of credential exposure, unauthorized network activity, or access to sensitive system directories was found. The skill operates within localized project directories.

baoyu-comic