skills/vltansky/skills/design-review/Gen Agent Trust Hub

design-review

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites via agent-browser to perform visual audits. This untrusted content could potentially contain instructions designed to influence the agent's subsequent actions, such as design recommendations or code fixes. \n
  • Ingestion points: Web content retrieved via agent-browser from user-specified TARGET_URL (SKILL.md). \n
  • Boundary markers: None explicitly defined for the audited web content. \n
  • Capability inventory: The skill can execute shell commands (git, grep) and write to the local filesystem to generate reports and apply UI fixes (SKILL.md). \n
  • Sanitization: No specific sanitization or filtering of the ingested web content is described before it is processed for design findings. \n- [COMMAND_EXECUTION]: The skill uses shell commands to manage its workflow and apply UI fixes. Commands include git status, git diff, git add, git commit, git revert, and grep (SKILL.md). These are used within the scope of its stated purpose but represent a significant capability tier. \n- [DYNAMIC_EXECUTION]: The skill utilizes page.evaluate() within the agent-browser context to execute JavaScript on the target website. This is used to extract design-related metadata such as fonts, colors, and layout metrics (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 05:13 PM
Security Audit — agent-trust-hub — design-review