The Agent Skills Directory

[SAFE]: No malicious patterns such as prompt injection, obfuscation, or data exfiltration were detected. The skill's primary function is providing static design guidance.
[SAFE]: The upstream source 'vercel-labs/agent-skills' belongs to a trusted organization, and the skill content aligns with its stated purpose.
[SAFE]: The skill has a potential surface for indirect prompt injection because it reads user-provided files for design review. However, the risk is negligible as the skill possesses no dangerous capabilities such as network access or shell execution.
Ingestion point: User-provided design or code files (SKILL.md, Section 8.2).
Boundary markers: Absent.
Capability inventory: None. The skill only outputs text-based advice and configuration snippets.
Sanitization: Absent.
[SAFE]: All provided code snippets (CSS, Tailwind configuration, JavaScript) are standard configuration templates for user implementation and do not contain executable commands or backdoors.

web-design