The Agent Skills Directory

[SAFE]: The skill is well-structured and follows security best practices for AI agents, such as requiring user approval before executing generative tasks and using specific output directories to manage artifacts.
[COMMAND_EXECUTION]: The skill executes local scripts (scripts/merge-to-pdf.ts) and system utilities (pngquant, optipng, sips) for image processing and document generation. These commands are integral to the skill's primary function and are executed within the context of the user's project directory.
[PROMPT_INJECTION]: The skill processes external source content (Step 1.2), which presents a surface for indirect prompt injection. However, the risk is mitigated by the mandatory review steps where the agent presents the analysis and storyboard to the user for confirmation before proceeding to image generation.
Ingestion points: User-provided source content files or pasted text processed in analysis.md.
Boundary markers: None identified in the analysis or storyboard generation prompts.
Capability inventory: Subprocess execution via npx, file system access (read/write), and image generation.
Sanitization: The skill relies on standard model processing without specific character escaping or instruction filtering for the source content.

baoyu-comic