Skills
Navigate
skills/ai-hpc/claude-scholar/web-design-reviewer/Gen Agent Trust Hub

web-design-reviewer

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it navigates to external websites and retrieves their content while having permissions to write to the local source code repository.
  • Ingestion points: The skill utilizes browser_navigate to access remote URLs and browser_snapshot to retrieve DOM structures as described in SKILL.md.
  • Boundary markers: There are no explicit delimiters or protective instructions defined to prevent the agent from interpreting content on the scanned websites as instructions.
  • Capability inventory: The skill has powerful capabilities across its scripts, including browser_navigate, browser_take_screenshot, and File Read/Write for applying layout fixes.
  • Sanitization: There is no evidence of content filtering or validation for the data ingested from external web pages before it is used to determine code modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 03:42 PM