tathagat22

Shared distillation cache for AI agents — every fetch ~73-89% fewer tokens via a shared cache.

Two-way Figma MCP: extract + verify design-to-code, or generate on-brand Figma pages from a prompt.