The Agent Skills Directory

[SAFE]: The skill is designed with a strong focus on security, providing detailed instructions for implementing Row-Level Security (RLS) and data encryption using the pgcrypto extension.\n- [COMMAND_EXECUTION]: The provided validation script (scripts/validate_schema.py) is used for static analysis of SQL files. It does not invoke a shell, execute SQL code, or utilize any external dependencies, making it safe for local use.\n- [DATA_EXFILTRATION]: No network connectivity or data transmission logic is present in the skill's instructions or scripts. All operations are local.\n- [CREDENTIALS_UNSAFE]: The skill avoids hardcoded credentials and instead provides best-practice templates for using application-level settings and environment variables for sensitive parameters.\n- [PROMPT_INJECTION]: The instructions are purely technical and do not contain any attempts to bypass model safety guardrails or override system instructions.

database-design