service-objects-vs-fat-models
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and links to the Interactor gem and Dry-Monads documentation. These are established libraries within the Ruby on Rails development community and are considered safe sources.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze Ruby code provided by users, which introduces a vulnerability to indirect prompt injection.
- Ingestion points: The skill metadata and prompts indicate it is used to review and refactor controller actions and model files supplied by the user during the interaction (SKILL.md, evals.md).
- Boundary markers: The skill lacks explicit instructions to treat the analyzed code as untrusted or to use delimiters to prevent embedded commands from being executed.
- Capability inventory: The agent context typically includes access to file system operations and shell execution, which could be exploited via malicious code snippets if they are interpreted as instructions.
- Sanitization: No mechanisms for sanitizing or validating user-provided code strings are specified.
Audit Metadata