The Agent Skills Directory

[SAFE]: The skill demonstrates secure database interaction patterns by using parameter binding (e.g., :isActive) in the Query Builder examples. This is a primary defense against SQL injection attacks.
[SAFE]: The guidelines explicitly recommend setting synchronize: false for production environments. This prevents accidental schema modifications or data loss that can occur with TypeORM's auto-sync feature.
[SAFE]: Sensitive configuration values like database passwords and hostnames are shown being loaded from environment variables (process.env.DB_PASSWORD), which is the industry-standard practice for secure credential management.
[INDIRECT_PROMPT_INJECTION]: The skill defines an interface for processing database records.
Ingestion points: Data enters the system via userRepository find/save methods and createQueryBuilder operations defined in SKILL.md.
Boundary markers: The code uses TypeORM's named parameter syntax (:value) to separate instructions from data.
Capability inventory: The skill facilitates database read, write, and schema migration operations.
Sanitization: The implementation relies on TypeORM's underlying driver to handle escaping and sanitization of parameterized inputs.

typeorm