The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides a prompt for the agent to inspect a user-provided website URL and extract its design system. This represents an attack surface where a malicious website could host instructions designed to influence the agent's subsequent actions.
Ingestion points: External website content via [WEBSITE_URL] in the 'match an existing website' section of SKILL.md.
Boundary markers: The prompt uses some structure but lacks explicit delimiters or warnings for the agent to ignore instructions embedded in the external site content.
Capability inventory: The skill expects the agent to update configuration files (docs.config.ts), modify CSS, and potentially run a development server.
Sanitization: No content sanitization or validation steps are defined for the design token extraction process.
[EXTERNAL_DOWNLOADS]: The skill references various official packages and documentation resources.
References official NPM packages: @farming-labs/docs, @farming-labs/theme, @farming-labs/svelte-theme, @farming-labs/astro-theme, and @farming-labs/nuxt-theme.
Links to documentation and repositories at docs.farming-labs.dev and github.com/farming-labs/docs.
These are vendor-owned resources for the skill author and are considered safe.

creating-themes