The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides example command-line instructions for optimizing images using utilities like cwebp, mogrify, and jpegoptim. These are presented as guidance for the agent to help users with asset management.
[DATA_EXFILTRATION]: Includes a functional curl command example aimed at auditing image assets on a website (https://yoursite.com). While intended for SEO purposes, this demonstrates the capability to perform network operations to external, non-whitelisted domains.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read marketing context from local files (e.g., .agents/product-marketing.md), creating a surface for potential indirect prompt injection if those files contain untrusted content.
Ingestion points: The skill reads external context from files including .agents/product-marketing.md, .claude/product-marketing.md, and product-marketing-context.md (SKILL.md).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the ingestion logic.
Capability inventory: The skill utilizes command-line tools for file system interaction and network requests as part of its visual asset optimization and auditing functions.
Sanitization: There is no evidence of validation or sanitization of the content ingested from the marketing context files.
[EXTERNAL_DOWNLOADS]: Mentions the use of the @vercel/og Node.js package for dynamic image generation. This is an official package from a well-known service provider.

image