Yet-Another-Agentic-Web-Chat

YAAWC design system rules and tokens. MUST be used whenever editing or creating UI — any .tsx component, Tailwind class change, styling work, color/spacing/radius/shadow/typography decision, or globals.css edit. Activates on terms like "component", "style", "UI", "button", "card", "modal", "color", "theme", "dark mode", "Tailwind", "CSS", "layout", "padding", "spacing", "rounded", "shadow", "hover", "focus", "MessageBox", "Sidebar", "Navbar", "ChatWindow", or any work under src/components or src/app.

Use when modifying React UI components, ChatWindow, MessageBox, MarkdownRenderer, frontend state/streaming event handling, styling, or debugging UI rendering.

Build, run, and drive YAAWC (the Next.js AI search app in this repo). Use when asked to start YAAWC, launch the dev server, build it, smoke-test it, screenshot or inspect the home/dashboard UI, or interact with the running app in a browser.

Use when working on tool-call lifecycle events, todo updates, subagent streaming, SimplifiedAgent/ChatWindow event handling, MarkdownRenderer ToolCall, TodoWidget, or debugging missing or broken streaming output.

Use when adding, modifying, or debugging API endpoints, request/response schemas, payload formats, or the chat and search route handlers.

Use when working on image upload/display, the uploads API, MessageInput image/clipboard paste, multimodal chat messages, or vision model integration.

Use when adding or extending agent tools, LLM providers, API routes, or focus modes in YAAWC, or when asked how to add new functionality.

Use when modifying agent prompts, focus-mode prompts, citation formatting, personalization context injection, system prompts, or debugging how the agent responds.

Use when working on the deep_research tool, subagents, SubagentExecutor, SubagentExecution UI, subagent tool restrictions, or subagent event flow.

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

Use when testing or verifying changes, running the dev server, static analysis, curl API checks, playwright browser automation, or end-to-end validation.

This skill should be used when the user asks about libraries, frameworks, API references, or needs code examples. Activates for setup questions, code generation involving libraries, or mentions of specific frameworks like React, Vue, Next.js, Prisma, Supabase, etc.