The Agent Skills Directory

[SAFE]: The skill consistently promotes security best practices for database interactions.
[PROMPT_INJECTION]: No attempts to override agent behavior or bypass safety filters were detected. The instructions are focused on technical implementation details.
[DATA_EXFILTRATION]: No sensitive file access or unauthorized network operations were found. Database configuration examples correctly suggest using environment variables (e.g., DATABASE_URL).
[REMOTE_CODE_EXECUTION]: All code examples use well-known, reputable Go packages (pgx, squirrel, go-redis). No remote script execution or untrusted downloads are present.
[COMMAND_EXECUTION]: No shell commands or subprocess calls are executed within the skill instructions or scripts.
[DATA_EXPOSURE]: The skill emphasizes the use of parameterized queries and explicitly forbids string concatenation in SQL, effectively teaching how to prevent SQL injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: While the skill handles external data (database records), it provides clear boundaries and uses parameterization to ensure data is treated as values, not executable instructions.
Ingestion points: Database query parameters in SKILL.md and reference files.
Boundary markers: Explicit instructions to use $1 placeholders and library-specific builders (Squirrel).
Capability inventory: Database operations (CRUD) through pgxpool.
Sanitization: Native parameterization provided by the pgx driver.

go-db