The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. it ingests untrusted data from user-provided source files or pasted text (saved as source.md) to perform analysis and generate content. If this data contains malicious instructions, the agent may follow them during the storyboard or prompt generation phases. Ingestion points: source.md and user input. Boundary markers: Markdown code blocks. Capability inventory: File writing, script execution (merge-to-pdf), and external skill calls. Sanitization: No explicit sanitization or instructions to ignore embedded commands are present.
[COMMAND_EXECUTION]: The skill executes local TypeScript scripts using npx and bun. Specifically, it runs scripts/merge-to-pdf.ts to assemble the final PDF and calls a sibling skill script ../baoyu-image-gen/scripts/main.ts for image generation.
[EXTERNAL_DOWNLOADS]: The skill uses npx -y which can automatically download and install required Node.js packages from the official npm registry at runtime if they are missing from the environment.
[DATA_EXFILTRATION]: The skill accesses the user's home directory (~/.baoyu-skills/) to read and write persistent configuration files like EXTEND.md. While used for legitimate preference management, this involves accessing sensitive filesystem paths.

baoyu-comic