The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses database migrations and queries but correctly employs parameterized fragments and the pin operator (^) to prevent SQL injection. It explicitly documents string interpolation in fragments as a high-risk anti-pattern.
[DATA_EXFILTRATION]: There are no detected network operations or attempts to access sensitive system files. Database operations are scoped to the application's repository context.
[PROMPT_INJECTION]: The skill instructions focus on architectural patterns and code quality. No override or bypass patterns were detected in the instructions or metadata.
[REMOTE_CODE_EXECUTION]: No remote script downloads or dynamic execution patterns (like eval or Code.eval_string) are present in the provided files.
[SAFE]: The skill differentiates between cast (for external, untrusted input) and change/put_change (for internal, trusted data), which is a key security practice in Elixir applications to prevent mass assignment vulnerabilities.

ecto-patterns