The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines endpoints for user registration and login that process untrusted data, which represents a standard architectural attack surface for indirect injection if not combined with proper validation.
Ingestion points: Data enters the system via the @Body() decorator in the signIn and register controller methods in auth.controller.ts.
Boundary markers: The provided code snippets do not explicitly show the use of validation pipes or DTO decorators to enforce data boundaries, though this is common in NestJS implementations.
Capability inventory: The skill utilizes bcrypt for credential comparison and the JWT service for token generation and signing.
Sanitization: No explicit input sanitization is shown in the snippets, although the skill recommends implementing rate limiting via the ThrottlerModule to mitigate brute-force attacks.
[EXTERNAL_DOWNLOADS]: The skill instructs users to install standard, well-known authentication and security libraries from official registries.
Evidence: Instructions include npm install for @nestjs/jwt, passport, bcrypt, and related packages from the official NPM registry.

authentication