motion-review
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and serves to guide the AI agent in providing structured feedback on user interface motion code. It does not perform file operations or network requests.- [PROMPT_INJECTION]: The skill naturally processes untrusted code diffs as part of its core function, creating an indirect prompt injection surface. However, the skill lacks the necessary tool capabilities (such as shell access or network tools) to be exploited by malicious diff content. The risk is considered negligible given the skill's narrow scope.
- Ingestion points: The skill ingests and analyzes external code diffs provided in the conversation context.
- Boundary markers: None defined; the agent is instructed to read the entire diff.
- Capability inventory: Output is restricted to markdown-formatted review tables and text verdicts.
- Sanitization: No specific sanitization methods for input diffs are implemented.- [DATA_EXFILTRATION]: The skill does not access environment variables, credentials, or sensitive configuration files.- [REMOTE_CODE_EXECUTION]: There are no patterns involving the download or execution of remote scripts or unverified third-party packages.- [COMMAND_EXECUTION]: The skill mentions searching (grepping) for design tokens within the codebase, which is a standard and safe operation for a code-reviewing agent.
Audit Metadata