web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes both external content and local user files, creating a potential surface for indirect instructions to influence agent behavior.
- Ingestion points: Reads user-specified files and fetches remote guidelines from a GitHub URL.
- Boundary markers: The skill does not define specific delimiters or "ignore embedded instructions" warnings for the ingested content.
- Capability inventory: Capability to read local files and perform network requests (WebFetch).
- Sanitization: No explicit sanitization or validation of the fetched or read content is mentioned.
- [EXTERNAL_DOWNLOADS]: Fetches design guidelines from an external source during execution.
- Source: Downloads content from a public repository hosted on GitHub by a well-known organization.
- Context: The fetched content is used to define the rules for the design audit process.
Audit Metadata