The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to analyze local codebases for Tailwind and CSS configurations using fd and grep. It also includes a helper script, extract-palette.sh, which executes FFmpeg commands to extract color palettes from image files provided by the user.\n- [EXTERNAL_DOWNLOADS]: The skill fetches remote assets, including web manifests, stylesheets, and icons, from user-provided URLs using curl and WebFetch. It also recommends the installation and use of the node-vibrant package from the NPM registry for image processing tasks.\n- [PROMPT_INJECTION]: The skill processes untrusted content from external websites to analyze brand voice and typography, creating a risk for indirect prompt injection.\n
Ingestion points: Homepage hero text and deep-page copy fetched via WebFetch or curl (SKILL.md).\n
Boundary markers: No explicit delimitation or instructions are provided to the agent to treat fetched content as untrusted data.\n
Capability inventory: The skill has access to sensitive tools including Bash for command execution and Write/Edit for file system modification.\n
Sanitization: Fetched website content is not sanitized or filtered before it is used for analysis by the agent.

brand-kit