The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes the NPM registry to download and install dependencies and project templates through commands like pnpm add and pnpm dlx (e.g., nuxt-security, create-nuxt-base). These are standard operations for frontend development within the specified ecosystem.
[COMMAND_EXECUTION]: Instructions are provided for running local shell commands to manage the development lifecycle, such as starting development servers (pnpm run dev), generating API types (pnpm run generate-types), and executing End-to-End tests via Playwright.
[COMMAND_EXECUTION]: Includes administrative commands for environment maintenance, such as port management using lsof and kill to resolve process conflicts during development.
[PROMPT_INJECTION]: Indirect prompt injection surface identified. Ingestion points: Browser snapshots, network request logs, and local frontend code files. Boundary markers: None explicitly defined for agent instructions. Capability inventory: Shell command execution and browser interaction via Chrome DevTools MCP. Sanitization: Extensive guidance provided for sanitizing application data using DOMPurify and validating inputs with Valibot.

developing-lt-frontend