web-design-guidelines
Warn
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The skill's internal metadata claims the author is 'vercel', while the repository information and external metadata identify the author as 'sadroad'. This discrepancy is deceptive and misleads the user regarding the origin and trustworthiness of the design guidelines.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and follow instructions from external, potentially untrusted sources.
- Ingestion points: The agent is directed to read the 'command.md' file and various project files defined by a user-supplied pattern.
- Boundary markers: The instructions lack delimiters or constraints to prevent the agent from being influenced by malicious instructions embedded within the audited files or the guidelines file itself.
- Capability inventory: The skill utilizes file system read capabilities to ingest project content.
- Sanitization: There is no evidence of content sanitization or validation of the ingested files before they are processed by the agent.
Audit Metadata