web-design-reviewer
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to visit live websites and inspect their DOM structure using tools like
browser_navigateandbrowser_snapshot. This creates an attack surface where maliciously crafted web content could attempt to inject instructions into the agent's context. - Ingestion points: External URLs and DOM snapshots processed via browser MCP tools as described in
SKILL.md. - Boundary markers: Absent. The instructions do not define clear delimiters or warnings to ignore instructions embedded in the target web pages.
- Capability inventory: The agent has the capability to modify local source code and execute the provided
scripts/css-risk-audit.pyscript. - Sanitization: Absent. There is no mention of filtering or sanitizing content retrieved from the web before processing.
- [COMMAND_EXECUTION]: The skill includes and instructs the execution of a Python script (
scripts/css-risk-audit.py) which reads local project files to identify risky CSS patterns. - [EXTERNAL_DOWNLOADS]: The workflow requires the agent to navigate to external, live websites to perform visual QA and design reviews.
Audit Metadata