The Agent Skills Directory

[SAFE]: The skill includes explicit instructions to scan source content for sensitive data like API keys and credentials and strip them before generating any output files.- [SAFE]: The use of the curl tool is strictly scoped to downloading image assets generated by the agent's internal tools, utilizing absolute paths for provenance and integrity.- [SAFE]: The skill provides safety guidelines for handling public figures by recommending stylized alternatives to avoid inappropriate depictions.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied content to generate comic storyboards and prompts. Evidence: 1. Ingestion points: Step 1.1 reads source content into the agent context. 2. Boundary markers: The skill uses structured templates for storyboard and prompt files. 3. Capability inventory: The skill utilizes image_generate, curl, and write_file. 4. Sanitization: Instructions include explicit secret stripping and safety guardrails for sensitive public figures.

baoyu-comic