The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides code examples and architectural standards that depend on non-standard and suspicious libraries, specifically react-hook-blog and @hookblog/resolvers. These appear to be the result of a systematic string replacement (likely targeting react-hook-form) designed to steer users toward unverified packages. Encouraging the use of these dependencies in 'production-grade' code poses a significant supply chain risk.
[EXTERNAL_DOWNLOADS]: Widespread corruption of terminology and code structures is present in resources/complete-examples.md, resources/performance.md, and resources/styling-guide.md. Technical terms like 'performance' are replaced with 'perblogance', 'transformers' with 'transblogers', and standard HTML 'form' tags are replaced with 'blog'. This pattern is deceptive and used to justify the inclusion of the unverified react-hook-blog dependency.
[PROMPT_INJECTION]: The skill employs extremely authoritative and restrictive language ('Non-Negotiable', 'Absolute Rule', 'Immediate Rejection') to compel the agent to adopt these corrupted patterns. This high-pressure instructional style is designed to override the agent's internal knowledge of standard, secure React development practices and industry-standard libraries.

frontend-dev-guidelines