The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown documentation and code snippets intended to guide an AI agent or developer in writing and refactoring Ruby on Rails code. No executable scripts or configurations that could compromise the environment were found.
[EXTERNAL_DOWNLOADS]: The skill references several official Ruby on Rails and 37signals resources, including the Fizzy codebase and the Solid stack (Solid Queue, Solid Cache, Solid Cable). These are well-known, reputable open-source projects. There are no patterns of downloading and executing untrusted remote code (e.g., curl | bash).
[PROMPT_INJECTION]: There are no attempts to override system prompts or bypass safety filters. Instructional language like 'CRITICAL' or 'IMPORTANT' is used exclusively to categorize architectural priority and impact, which is benign.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or private tokens were detected. Code examples demonstrating authentication use standard security practices like has_secure_password and token generation with placeholders.
[DATA_EXFILTRATION]: No network operations or file system access patterns targeting sensitive user data or system configuration files were identified.

37signals-rails