design-review
Warn
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructions specify the use of
npx saglitzdesign-mcpto access rubrics and auditing tools. This command performs the immediate execution of a remote package downloaded from the npm registry. \n- [EXTERNAL_DOWNLOADS]: The skill depends on fetching the latest version of an external Node.js package from the public npm repository at runtime to perform its design audits. \n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external, untrusted content such as live web pages and source code. This represents an attack surface where malicious instructions could be embedded in the analyzed UI to influence agent behavior. \n - Ingestion points: The skill explicitly processes screenshots, live pages, and code snippets from external sources. \n
- Boundary markers: There are no instructions provided to delimit external content or to ignore potential instructions within that content. \n
- Capability inventory: The skill uses shell execution capabilities (
npx) and several specific design tools that process data. \n - Sanitization: No sanitization or validation steps are defined for the ingested UI data.
Audit Metadata