The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses bun to execute a local script (.agents/skills/baoyu-image-gen/scripts/main.ts) for image generation. The command line arguments, including the output paths and filenames (slugs), are derived from user-provided article content. While the instructions specify that slugs should be in 'kebab-case', there is a risk of command injection if the input is not strictly validated before being passed to the shell command.\n- [PROMPT_INJECTION]: The skill processes untrusted article content to generate visual outlines and image prompts, creating an indirect prompt injection surface. Malicious instructions within an article could attempt to override the agent's analysis or influence the generated prompts. The workflow includes a 'Confirm Settings' step with a user question, which acts as a manual guardrail to mitigate this risk.\n- [DATA_EXPOSURE]: The skill instructions specify relying on .agents/skills/.env for model and provider settings during image generation. While this is a standard practice for managing API keys in development environments, it highlights an access point for sensitive configuration data that the skill's scripts interact with.

baoyu-article-illustrator