web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill fetches a remote Markdown file (
command.md) fromvercel-labs/web-interface-guidelines. Per the [TRUST-SCOPE-RULE], this is rated LOW as the organization is on the trusted list. - PROMPT_INJECTION (LOW): The skill implements an indirect prompt injection surface (Category 8). It retrieves instructions ("rules and output format instructions") from an external URL and applies them directly to the agent's logic.
- Ingestion points:
SKILL.mdtriggers a WebFetch of the guidelines URL. - Boundary markers: Absent; the agent is directed to follow the fetched content without delimiters.
- Capability inventory: Reading local files (specified by user) and generating formatted output.
- Sanitization: Absent; the skill relies entirely on the integrity of the remote GitHub content.
Audit Metadata