The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation file references/premium-patterns/09-tabs-image-swap/description.md references an external URL mivora-wbs.framer.website in its frontmatter. Automated security scanners have confirmed that this specific domain is a phishing site designed to deceive users.
[DATA_EXFILTRATION]: The inclusion of a confirmed phishing domain within the skill's reference materials poses a severe risk of credential theft and data exfiltration. The file references/premium-patterns/09-tabs-image-swap/description.md has been explicitly flagged as malicious by file reputation scanners due to its content.
[REMOTE_CODE_EXECUTION]: The skill instructs the agent to install various Node.js dependencies (e.g., framer-motion, lucide-react) and use tools like npx shadcn@latest to add components. It also references external components from the 21st.dev platform, creating a supply chain attack surface through unverifiable third-party packages and code providers.
[INDIRECT_PROMPT_INJECTION]: The skill contains a significant attack surface for indirect prompt injection.
Ingestion points: The skill processes user-supplied briefs and content within SKILL.md to generate web pages.
Boundary markers: Absent. There are no clear delimiters or instructions to ignore embedded commands in the processed data.
Capability inventory: The skill uses tools with Read, Write, Edit, Glob, and Grep permissions, and performs network operations via the fuse-browser MCP (referenced in references/design-inspiration.md).
Sanitization: There is no evidence of sanitization or filtering for external content before it is interpolated into prompts or written to files.

design-web