web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill fetches content from
https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md. While fetching remote instructions is a potential risk vector, the source is a verified trusted organization (vercel-labs), resulting in a downgrade to LOW per the [TRUST-SCOPE-RULE]. - PROMPT_INJECTION (LOW): The skill uses
WebFetchto retrieve a Markdown file containing 'rules and output format instructions' that guide the agent's behavior. This represents an Indirect Prompt Injection surface (Category 8). - Ingestion points: Remote URL (guidelines) and local filesystem (user UI code).
- Boundary markers: None specified in the skill definition to isolate the fetched guidelines from the agent's core instructions.
- Capability inventory: Reading local files, fetching web content, and generating text reports.
- Sanitization: No explicit sanitization of the fetched guidelines is performed. However, as the source is trusted and the skill's capabilities are limited to 'display only' (producing a report), the severity is low.
Audit Metadata