design-system-maintenance
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of application source code, component properties, and network request/response bodies during UI audits and migration parity checks.
- Ingestion points: The skill reads application code, properties, styles, and stories in COMPONENT-DOCS.md, and network request/response bodies and payloads in MIGRATION.md.
- Boundary markers: No explicit delimiters or instructions are specified to prevent the agent from being influenced by malicious instructions embedded in the processed application data.
- Capability inventory: The skill employs browser automation via Playwright and several JavaScript scripts (e.g., pixdiff.mjs, capture.mjs, bodies-check.mjs) for file analysis, UI rendering, and network monitoring.
- Sanitization: The instructions do not specify any validation or sanitization of the external content before processing.
- [EXTERNAL_DOWNLOADS]: The skill refers to external scripts and other skills hosted on the author's public GitHub repository.
- Evidence: References to github.com/Quegenx/agent-skills are used for design token definitions, accessibility requirements, and utility scripts like pixdiff.mjs and capture-guard.mjs. These are vendor-owned resources relevant to the skill's function.
- [COMMAND_EXECUTION]: The skill's instructions involve the use of browser automation and analysis scripts.
- Evidence: The workflow includes using Playwright to capture UI snapshots and intercept network traffic, and executing scripts for image comparison and data validation. This behavior is expected for a UI migration and design auditing tool.
Audit Metadata