The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill promotes secure credential management by requiring authentication tokens to be read from environment variables rather than being hardcoded in source files.\n- [PROMPT_INJECTION]: Documentation explicitly warns against SQL injection and provides correct usage patterns for parameterized queries to ensure untrusted input is not executed as code.\n- [COMMAND_EXECUTION]: The provided scripts/verify.sh is a local utility for static analysis that does not perform network operations or execute remote code, focusing instead on detecting configuration errors like hardcoded tokens.\n- [SAFE]: All external references and dependencies, such as the @libsql/client library, are standard for the technology and used in a manner consistent with official documentation and safety guidelines.

sqlite-turso