The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides patterns for ingesting and rendering external or user-provided data, such as markdown processing, list filtering, and data processing via Web Workers. This represents a standard application attack surface where malicious input could influence the agent's behavior if not properly handled.
Ingestion points: Raw data processing in resources/data-and-lists.md (Web Workers), user input for filtering in resources/rendering.md, and markdown ingestion in resources/bundle.md.
Boundary markers: The instructions do not explicitly mandate specific boundary markers for all data ingestion scenarios, though they provide general architectural advice.
Capability inventory: The skill does not provide dangerous capabilities like arbitrary shell execution or file system write access combined with this data ingestion.
Sanitization: The skill explicitly recommends security best practices, such as using DOMPurify to sanitize rendered markdown in resources/bundle.md.
[EXTERNAL_DOWNLOADS]: The skill recommends several well-known and standard Node.js packages for performance optimization and measurement. These are industry-standard tools.
Evidence: Recommendations for web-vitals, react-window, react-virtuoso, and official React/Next.js tooling such as babel-plugin-react-compiler.

react-performance