Skills
Navigate
skills/github/awesome-copilot/web-design-reviewer/Gen Agent Trust Hub

web-design-reviewer

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes untrusted content from user-provided URLs, which can be used to influence agent behavior through indirect instructions embedded in web pages.
  • Ingestion points: The skill uses browser_navigate and browser_snapshot to ingest DOM structure and visual content from external websites (SKILL.md).
  • Boundary markers: The instructions lack explicit boundary markers or directives for the agent to ignore instructions found within the target website's content.
  • Capability inventory: The agent has permission to perform File Read/Write operations, Code Search, and grep_search across the local workspace to apply fixes (SKILL.md).
  • Sanitization: No sanitization or validation of the external web data is mentioned before it is used to determine source code modifications.
  • [EXTERNAL_DOWNLOADS]: The skill's implementation guidance recommends downloading the @playwright/mcp package from the npm registry.
  • Evidence: The configuration example suggests using npx -y @playwright/mcp@latest (SKILL.md).
  • [REMOTE_CODE_EXECUTION]: The skill instructions describe executing a remote package to provide browser automation capabilities.
  • Evidence: The 'Reference Implementation' section provides a command to run the @playwright/mcp server using npx (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 08:22 PM
Security Audit — agent-trust-hub — web-design-reviewer