The Agent Skills Directory

[COMMAND_EXECUTION]: The setup script executes standard system and developer tools to scaffold the application and manage dependencies. It uses npx for project creation, npm for package management, and openssl for secret generation.
Evidence: scripts/setup-frontend.sh executes npx create-next-app and npm install for required dependencies.
[INDIRECT_PROMPT_INJECTION]: The generated Next.js application fetches article data from an external Drupal backend and renders it in the UI. This creates a surface where malicious instructions embedded in article content could potentially influence the behavior of an AI agent observing the rendered application.
Ingestion points: assets/app/lib/drupal.ts (the fetchArticles function fetches data from the Drupal JSON:API).
Boundary markers: No specific delimiters or "ignore instructions" markers are present in the rendered output within assets/app/app/page.tsx.
Capability inventory: The skill possesses capabilities for shell script execution (scripts/setup-frontend.sh), file system writes, and local network operations.
Sanitization: The application uses standard React JSX rendering which provides protection against XSS, but it does not implement specific filtering for natural language instructions within the fetched article titles.

frontend-connect