The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied articles or stories to generate creative content such as storyboards and image prompts. This creates an attack surface where maliciously crafted input could influence the agent's behavior during the generation process.
Ingestion points: Source material is read from user input or files and saved to source.md (e.g., in SKILL.md Step 1.2 and references/workflow.md).
Boundary markers: The workflow lacks explicit delimiters or instructions to the agent to ignore potentially malicious embedded commands within the user-provided source content during analysis and prompt generation.
Capability inventory: The skill utilizes file system operations (read/write) and shell command execution via npx (for image generation and PDF assembly) across multiple scripts and workflow steps.
Sanitization: There is no evidence of sanitization or filtering applied to the ingested content before it is processed by the agent.

baoyu-comic