The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill provides a code example for internationalization in references/patterns.md that uses dynamic imports with template literals (await import(../messages/${locale}.json)). This pattern facilitates loading modules from computed paths, which can lead to unauthorized file access or directory traversal if the locale variable is derived from untrusted input without validation.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data for email template generation, creating an attack surface for indirect prompt injection.
Ingestion points: The skill processes user-supplied template properties (e.g., WelcomeEmailProps), internationalization JSON files, and brand configuration assets like logo URLs in SKILL.md and references/patterns.md.
Boundary markers: The instructions recommend avoiding ESP placeholders like {{mustache}} in JSX but do not provide the agent with specific delimiters or instructions to treat interpolated data as untrusted content.
Capability inventory: The skill utilizes capabilities to render templates to HTML and plain text, export files to the local file system, and send emails via external service integrations (Resend, SendGrid, Nodemailer) as detailed in references/render-send.md.
Sanitization: While references/components.md cautions that the Markdown component should only be used for trusted strings, there are no broader instructions for sanitizing or escaping properties before interpolation.
[EXTERNAL_DOWNLOADS]: The skill references and provides instructions for installing numerous external packages from npm, including react-email, @react-email/editor, and @react-email/render, and points to documentation hosted on react.email and GitHub.

react-email