The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions require the agent to ingest data from external context files, creating a potential surface for indirect prompt injection. \n
Ingestion points: The agent is directed to read .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md for brand and product details as specified in SKILL.md. \n
Boundary markers: There are no explicit instructions for the agent to treat the contents of these files as untrusted or to ignore embedded instructions. \n
Capability inventory: SKILL.md contains example shell commands for image optimization and inspection (e.g., cwebp, mogrify, jpegoptim, curl). \n
Sanitization: No sanitization or validation steps are provided for the content read from these external files. \n- [EXTERNAL_DOWNLOADS]: The skill identifies and recommends many well-known and trusted external services for design and AI image generation, such as Google Gemini, OpenAI, Vercel, Cloudflare, Canva, and Figma. These references are appropriate for the skill's purpose. \n- [COMMAND_EXECUTION]: The skill includes documentation for several shell commands used for image optimization and web page inspection. These are provided as educational examples for the user to execute manually and are not designed for automated execution by the agent.

image