The Agent Skills Directory

[SAFE]: The skill documents internal security protocols for a specific administrative framework. It mandates three layers of guards: UI-level route checks, server-side middleware for RPC calls, and database-level access control.
[SAFE]: Administrative database access is handled via a dedicated environment variable (LAT_ADMIN_DATABASE_URL), which is a standard practice for managing secrets in the provided architecture. The skill provides instructions on bypassing Row Level Security (RLS) only when using this specific administrative context, which is the intended purpose of the backoffice tools.
[SAFE]: The skill emphasizes 'Security by Obscurity' by requiring NotFoundError instead of Unauthorized responses, preventing non-admin users from fingerprinting the existence of administrative endpoints.
[SAFE]: Impersonation logic is described with specific middleware requirements (impersonatingMiddleware) and audit logging, ensuring that these high-privilege operations are gated and monitored according to the system's design.
[SAFE]: No malicious patterns such as prompt injection, data exfiltration, or unauthorized remote code execution were detected. The skill's instructions focus on maintaining the integrity and confidentiality of the administrative interface.

backoffice