The Agent Skills Directory

[PROMPT_INJECTION]: No behavior-override instructions or safety bypass attempts were detected in the skill markdown or scripts.\n- [DATA_EXFILTRATION]: The skill does not perform network operations, and the tools strictly process local inputs provided by the user. No hardcoded credentials or sensitive file paths are accessed.\n- [OBFUSCATION]: The skill content is clear and readable text, with no evidence of Base64 encoding, zero-width characters, or homoglyph attacks.\n- [REMOTE_CODE_EXECUTION]: The skill provides its own logic and does not download or execute remote scripts or external packages.\n- [PRIVILEGE_ESCALATION]: The skill does not use sudo, chmod 777, or other mechanisms to acquire elevated permissions.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted SQL DDL and JSON data. However, risk is mitigated as the schema analyzer strips SQL comments, and the skill lacks network access or high-risk execution capabilities. Ingestion points include sample and user-provided asset files.

database-designer