The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents how to execute database commands safely. It contains multiple 'Critical Requirements' and 'Red Flags' specifically instructing the agent to use prepared statements (execute()) with ? placeholders for all user-provided data, providing robust protection against SQL injection attacks.
[CREDENTIALS_UNSAFE]: The skill follows security best practices for secret management by recommending the use of environment variables (process.env.DATABASE_URL). While a snippet contains a hardcoded password example, it is explicitly labeled as a 'Bad Example' to demonstrate insecure practices that must be avoided.
[INDIRECT_PROMPT_INJECTION]: The skill addresses potential injection risks arising from processing untrusted database content by enforcing strict parameterization for all queries. It provides clear boundary markers to ensure that data is treated as values rather than executable SQL code.
[SAFE]: The skill is purely informative and pedagogical, providing legitimate development patterns without any hidden instructions, obfuscation, or unauthorized data exfiltration attempts. It focuses on stable, well-known libraries and standard Node.js APIs.

api-database-mysql