The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions and documentation are purely technical and do not contain any instructions attempting to override agent behavior, bypass safety filters, or extract system prompts.
[CREDENTIALS_UNSAFE]: The snippets include example environment variables such as API_SECRET_KEY=sk_live_xxxxx. These are clearly identified as placeholders in a documentation context and do not expose actual secrets.
[EXTERNAL_DOWNLOADS]: No unauthorized external downloads or remote scripts are triggered. The snippets reference well-known, legitimate libraries like @tanstack/react-query, zod, and next.
[COMMAND_EXECUTION]: The provided code focuses on application-level logic (API routes, UI components) and does not involve the execution of shell commands, subprocess spawning, or system-level configuration changes.
[DATA_EXFILTRATION]: Network operations described (fetch, GraphQL, WebSockets) are standard for fullstack applications and use placeholder domains (example.com). There is no logic present that reads sensitive system files for exfiltration.
[SAFE]: The skill demonstrates defensive coding practices, such as using Zod for build-time environment variable validation and implementing token checks for preview modes.

fullstack-modern