The Agent Skills Directory

[SAFE]: The skill provides documentation and SQL templates for standard database operations. No malicious patterns or prompt injection attempts were detected in SKILL.md or the reference guides.\n- [SAFE]: The script scripts/validate.py and the configuration patterns in references/PATTERNS.md employ yaml.safe_load(), which is the industry standard for preventing unsafe deserialization and arbitrary code execution vulnerabilities during YAML processing.\n- [SAFE]: All metadata fields, including authors and bonded agent references, are consistent with the skill's stated purpose and the provided file structure, with no evidence of metadata poisoning or deceptive techniques.

postgresql-json