Demos
715 resultsMost Popular
CopilotKit's Built-in AgentCLI Start Command
Copy-paste command to clone the canonical starter
CopilotKit's Built-in AgentBeautiful Chat
Polished starter chat with two-pane layout and pre-wired suggestions
CopilotKit's Built-in AgentAgentic Chat
Natural conversation with frontend tool execution
CopilotKit's Built-in AgentPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
CopilotKit's Built-in AgentPre-Built: Popup
Floating popup chat via <CopilotPopup />
CopilotKit's Built-in AgentChat Customization (Slots)
Customize CopilotChat via its slot system
CopilotKit's Built-in AgentChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
CopilotKit's Built-in AgentHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
CopilotKit's Built-in AgentHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
CopilotKit's Built-in AgentIn-Chat Human in the Loop
User approves agent actions before execution
CopilotKit's Built-in AgentTool Rendering
Backend agent tools rendered as UI components
CopilotKit's Built-in AgentTool-Based Generative UI
Agent uses tools to trigger UI generation
CopilotKit's Built-in AgentAgentic Generative UI
Long-running agent tasks with generated UI
CopilotKit's Built-in AgentShared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back
CopilotKit's Built-in AgentState Streaming
Per-token state delta streaming from agent to UI
CopilotKit's Built-in AgentSub-Agents
Multiple agents with visible task delegation
CopilotKit's Built-in AgentIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
CopilotKit's Built-in AgentMCP Apps
MCP server-driven UI via activity renderers
CopilotKit's Built-in AgentFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
CopilotKit's Built-in AgentOpen-Ended Generative UI (Advanced)
Sandboxed iframe UI calls back into host-defined functions
CopilotKit's Built-in AgentTool Rendering: Default Catch-all
Built-in DefaultToolCallRenderer wired with no config
CopilotKit's Built-in AgentTool Rendering: Custom Catch-all
Single branded wildcard renderer for every tool call
CopilotKit's Built-in AgentFrontend Tools
Tool defined in React, executed in the browser, invoked by the agent
CopilotKit's Built-in AgentFrontend Tools (Async)
Async frontend-tool handler with branded per-tool render
CopilotKit's Built-in AgentIn-App Human in the Loop
Async frontend-tool handler resolved by an app-level approval modal
CopilotKit's Built-in AgentReadonly Agent Context
Frontend publishes structured read-only context to the agent via useAgentContext
CopilotKit's Built-in AgentAuthentication
Bearer-token gate on the runtime via the V2 onRequest hook
CopilotKit's Built-in AgentAgent Config Object
Typed config (tone / expertise / responseLength) forwarded via provider properties
CopilotKit's Built-in AgentVoice Input
Speech-to-text via @copilotkit/voice with sample-audio button
CopilotKit's Built-in AgentMultimodal Attachments
Image and PDF uploads answered by the vision-capable gpt-4o agent
CopilotKit's Built-in AgentBYOC: Hashbrown
Streaming structured output via @hashbrownai/react
CopilotKit's Built-in AgentBYOC: json-render
Streaming structured output via @json-render/react
CopilotKit's Built-in AgentAgentic Chat (Reasoning)
Visible chain-of-thought during normal conversation, rendered via a custom reasoningMessage slot
CopilotKit's Built-in AgentReasoning (Default Render)
Zero-config rendering of the agent's reasoning chain via CopilotKit's built-in reasoning message
CopilotKit's Built-in AgentTool Rendering: Reasoning Chain
Sequential tool calls interleaved with the agent's visible reasoning chain
CopilotKit's Built-in AgentDeclarative Generative UI (A2UI Dynamic)
Agent designs an A2UI component tree at runtime from a registered catalog
CopilotKit's Built-in AgentA2UI Fixed Schema (Flight Card)
Frontend owns the component tree; the agent only streams data
CopilotKit's Built-in AgentGen UI Interrupt (Frontend Tool)
In-chat time-picker via useFrontendTool with an async handler that blocks until the user picks a slot — Strategy B adaptation of the LangGraph interrupt primitive
CopilotKit's Built-in AgentHeadless Interrupt (Frontend Tool)
Time-picker popup rendered outside the chat in the app surface — headless variant of the Strategy B interrupt adaptation
LangGraph (Python)LangGraph PlatformCLI Start Command
Copy-paste command to clone the canonical starter
LangGraph (Python)LangGraph PlatformPre-Built: CopilotChat
Vanilla CopilotChat with three starter-prompt suggestions — the minimum-viable surface for free-form chat
LangGraph (Python)LangGraph PlatformChat Customization: CSS
Default CopilotChat re-themed via CopilotKitCSSProperties
LangGraph (Python)LangGraph PlatformGenerative UI: Tool Rendering (Default)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
LangGraph (Python)LangGraph PlatformGenerative UI: Tool Rendering (Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
LangGraph (Python)LangGraph PlatformFrontend Tools: In-app Actions
Agent invokes client-side handlers registered with useFrontendTool
LangGraph (Python)LangGraph PlatformFrontend Tools: Async
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
LangGraph (Python)LangGraph PlatformHuman In the Loop: In-chat
Time-slot picker rendered inline in the chat via `useHumanInTheLoop` — the agent requests user input, the frontend renders a TimePickerCard, and the user's choice resumes the agent.
LangGraph (Python)LangGraph PlatformHuman in the Loop: In-app
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
LangGraph (Python)LangGraph PlatformGenerative UI: Tool Rendering (Custom)
Custom per-tool renderers (WeatherCard, FlightListCard) plus a wildcard catch-all for every other tool
LangGraph (Python)LangGraph PlatformGenerative UI: Tool calls + reasoning
Per-tool renderers (WeatherCard, FlightListCard, custom catchall) interleaved with a reasoningMessage slot rendering the agent's reasoning tokens — combines reasoning-display + tool-rendering in one chat surface.
LangGraph (Python)LangGraph PlatformGenerative UI: useComponent
Agent uses tools to trigger UI generation
LangGraph (Python)LangGraph PlatformGenerative UI: Agent State
Agent-state-driven Gen UI — the agent plans a live step list via Command(update={steps}); the frontend subscribes through useAgent and renders the steps inside CopilotChat's messageView.children slot.
LangGraph (Python)LangGraph PlatformShared State: Streaming
Per-token state delta streaming from agent to UI
LangGraph (Python)LangGraph PlatformShared State: Read + Write
Bidirectional shared state — UI writes preferences via agent.setState; agent writes notes via a Command-returning tool
LangGraph (Python)LangGraph PlatformShared State: Read-only
Recipe editor publishes form state via agent.setState; the agent reads the recipe context but does not mutate it (no backend tool — neutral default agent).
LangGraph (Python)LangGraph PlatformSub-Agents
Multiple agents with visible task delegation
LangGraph (Python)LangGraph PlatformPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
LangGraph (Python)LangGraph PlatformPre-Built: Popup
Floating popup chat via <CopilotPopup />
LangGraph (Python)LangGraph PlatformChat Customization: Slots
Customize CopilotChat via its slot system
LangGraph (Python)LangGraph PlatformHeadless UI: Simple
Minimum viable headless chat — useAgent + useCopilotKit dressed in shadcn/ui primitives. No tool rendering, no generative UI, just text in / text out.
LangGraph (Python)LangGraph PlatformHeadless UI: Complete
Full headless surface — hand-rolled CopilotChat replacement that wires every render hook (useRenderTool, useDefaultRenderTool, useComponent, useRenderToolCall, useRenderActivityMessage, useSuggestions, useAttachments) on top of shadcn primitives.
LangGraph (Python)LangGraph PlatformReasoning: Default
Built-in CopilotChatReasoningMessage rendering with no slot override.
LangGraph (Python)LangGraph PlatformReasoning: Custom
Visible reasoning/thinking chain alongside the final answer
LangGraph (Python)LangGraph PlatformHuman in the Loop: Interrupts
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the LangGraph interrupt lifecycle
LangGraph (Python)LangGraph PlatformHuman in the Loop: Headless Interrupts
Same interrupt(...) backend pattern as gen-ui-interrupt, but the time-picker mounts in a separate app-surface pane (left) instead of inline inside the chat — uses useHeadlessInterrupt (custom-event subscribe + manual runAgent forwardedProps.command.resume).
LangGraph (Python)LangGraph PlatformDeclarative UI: Dynamic A2UI
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; runtime injects the render_a2ui tool automatically
LangGraph (Python)LangGraph PlatformDeclarative UI: Fixed A2UI
Render an A2UI tree from a fixed, server-side schema — the agent streams data into a pre-authored component tree
LangGraph (Python)LangGraph PlatformMCP Apps
MCP server-driven UI via activity renderers
LangGraph (Python)LangGraph PlatformShared State: Frontend Context
Frontend provides read-only context to the agent via useAgentContext
LangGraph (Python)LangGraph PlatformDeclarative UI: json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
LangGraph (Python)LangGraph PlatformBeautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
LangGraph (Python)LangGraph PlatformAttachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
LangGraph (Python)LangGraph PlatformAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
LangGraph (Python)LangGraph PlatformDeclarative UI: Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
LangGraph (Python)LangGraph PlatformOpen Generative UI: Default
Agent composes UI from a registered component library — distinct from Tool Rendering, which attaches a custom renderer to a *named* backend tool. Use Open Generative UI when you want the agent to assemble UI freely from a palette you control.
LangGraph (Python)LangGraph PlatformOpen Generative UI: Custom
Same Open Generative UI surface, with the iframe-rendered components able to call back into frontend sandbox functions (e.g., a Calculator app whose buttons trigger handlers in the host page).
LangGraph (Python)LangGraph PlatformVoice
Speech-to-text via @copilotkit/voice with a bundled sample audio button
LangGraph (Python)LangGraph PlatformAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
LangGraph (TypeScript)LangGraph PlatformCLI Start Command
Copy-paste command to clone the canonical starter
LangGraph (TypeScript)LangGraph PlatformAgentic Chat
Natural conversation with frontend tool execution
LangGraph (TypeScript)LangGraph PlatformIn-Chat Human in the Loop (Original)
User approves agent actions before execution
LangGraph (TypeScript)LangGraph PlatformIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
LangGraph (TypeScript)LangGraph PlatformIn-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
LangGraph (TypeScript)LangGraph PlatformIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
LangGraph (TypeScript)LangGraph PlatformTool Rendering
Backend agent tools rendered as UI components
LangGraph (TypeScript)LangGraph PlatformTool-Based Generative UI
Agent uses tools to trigger UI generation
LangGraph (TypeScript)LangGraph PlatformAgentic Generative UI
Long-running agent tasks with generated UI
LangGraph (TypeScript)LangGraph PlatformState Streaming
Per-token state delta streaming from agent to UI
LangGraph (TypeScript)LangGraph PlatformShared State (Read + Write)
Bidirectional shared state — UI writes preferences via agent.setState; agent writes notes via a Command-returning tool
LangGraph (TypeScript)LangGraph PlatformSub-Agents
Supervisor delegates to research / writing / critique sub-agents exposed as tools — each delegation appended to a live log in shared state
LangGraph (TypeScript)LangGraph PlatformBeautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
LangGraph (TypeScript)LangGraph PlatformPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
LangGraph (TypeScript)LangGraph PlatformPre-Built: Popup
Floating popup chat via <CopilotPopup />
LangGraph (TypeScript)LangGraph PlatformChat Customization (Slots)
Customize CopilotChat via its slot system
LangGraph (TypeScript)LangGraph PlatformChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
LangGraph (TypeScript)LangGraph PlatformHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
LangGraph (TypeScript)LangGraph PlatformHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
LangGraph (TypeScript)LangGraph PlatformAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
LangGraph (TypeScript)LangGraph PlatformMultimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
LangGraph (TypeScript)LangGraph PlatformAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
LangGraph (TypeScript)LangGraph PlatformTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
LangGraph (TypeScript)LangGraph PlatformTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
LangGraph (TypeScript)LangGraph PlatformTool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
LangGraph (TypeScript)LangGraph PlatformReasoning
Visible reasoning/thinking chain alongside the final answer
LangGraph (TypeScript)LangGraph PlatformReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
LangGraph (TypeScript)LangGraph PlatformIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the LangGraph interrupt lifecycle
LangGraph (TypeScript)LangGraph PlatformHeadless Interrupt (testing)
Resolve langgraph interrupts from a plain button grid — no chat, no useInterrupt render prop
LangGraph (TypeScript)LangGraph PlatformDeclarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; backend agent owns the generate_a2ui tool
LangGraph (TypeScript)LangGraph PlatformDeclarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
LangGraph (TypeScript)LangGraph PlatformMCP Apps
MCP server-driven UI via activity renderers
LangGraph (TypeScript)LangGraph PlatformFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
LangGraph (TypeScript)LangGraph PlatformFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
LangGraph (TypeScript)LangGraph PlatformReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
LangGraph (TypeScript)LangGraph PlatformVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
LangGraph (TypeScript)LangGraph PlatformBYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
LangGraph (TypeScript)LangGraph PlatformBYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
LangGraph (TypeScript)LangGraph PlatformFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
LangGraph (TypeScript)LangGraph PlatformOpen-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
MastraMastra CloudCLI Start Command
Copy-paste command to scaffold the canonical Mastra starter
MastraMastra CloudAgentic Chat
Natural conversation with frontend tool execution
MastraMastra CloudIn-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
MastraMastra CloudIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat
MastraMastra CloudTool Rendering
Custom per-tool renderers plus a wildcard catch-all for every other tool
MastraMastra CloudTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines tools; frontend adds zero custom renderers
MastraMastra CloudTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool
MastraMastra CloudAgentic Generative UI
Long-running agent tasks with generated UI
MastraMastra CloudState Streaming
Per-token state delta streaming from agent to UI
MastraMastra CloudReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
MastraMastra CloudPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
MastraMastra CloudPre-Built: Popup
Floating popup chat via <CopilotPopup />
MastraMastra CloudChat Customization (Slots)
Customize CopilotChat via its slot system
MastraMastra CloudChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
MastraMastra CloudHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
MastraMastra CloudFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
MastraMastra CloudFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation and uses the returned result
MastraMastra CloudReasoning
Visible reasoning/thinking chain alongside the final answer (via custom reasoningMessage slot)
MastraMastra CloudReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
MastraMastra CloudAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent via useAgentContext
MastraMastra CloudDeclarative Generative UI (A2UI — Dynamic Schema)
Dynamic A2UI surfaces generated by the `generate_a2ui` tool; frontend supplies a typed catalog (Card, Metric, PieChart, BarChart, …)
MastraMastra CloudA2UI Fixed Schema
Frontend-owned component tree; agent streams data into the data model (flight card built from Card > Column > Title/Airport/Arrow/AirlineBadge/PriceTag/Button)
MastraMastra CloudHeadless Chat (Complete)
Full chat-from-scratch on useAgent — custom bubbles, manual generative-UI composition via useRenderToolCall / useRenderActivityMessage / useRenderCustomMessages
MastraMastra CloudAuthentication
Bearer-token gate enforced by the V2 runtime's onRequest hook — unauthenticated requests 401 before reaching the agent
MastraMastra CloudShared State (Read + Write)
Bidirectional shared state — UI writes preferences via agent.setState; backend reads them every turn and writes notes back via the set_notes tool
MastraMastra CloudSub-Agents
Supervisor Mastra Agent delegates to research / writing / critique sub-agents; every delegation appended to a live UI log via shared state
MastraMastra CloudBeautiful Chat (Flagship)
Flagship CopilotKit showcase combining A2UI, Open Generative UI, and MCP Apps in one cell
MastraMastra CloudBYOC — Hashbrown
Bring-your-own-component dashboard rendered via @hashbrownai/react useUiKit + useJsonParser
MastraMastra CloudBYOC — JSON Render
Bring-your-own-component dashboard rendered from a streamed JSON envelope into a frontend-owned catalog
MastraMastra CloudMultimodal Attachments
Image + PDF uploads forwarded to a vision-capable Mastra agent (gpt-4o)
MastraMastra CloudOpen Generative UI
Agent-authored HTML + CSS streamed into a sandboxed iframe via the runtime's openGenerativeUI flag
MastraMastra CloudOpen Generative UI (Advanced)
Open Generative UI with host-side sandbox functions wired through the iframe bridge
MastraMastra CloudVoice Input
Mic-to-text input via OpenAI Whisper, streamed into the chat composer
MastraMastra CloudTool-Based Generative UI
Agent uses a frontend tool to trigger rich generative UI inline in the chat
MastraMastra CloudHuman in the Loop (Step Selection)
Step-selection HITL surface rendered inline via `useHumanInTheLoop` (the LangGraph-specific `useLangGraphInterrupt` branch is a no-op on Mastra)
MastraMastra CloudIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow (alias of hitl-in-chat)
MastraMastra CloudMCP Apps
MCP server-driven UI rendered inline via the runtime's mcpApps config and the built-in activity renderer
MastraMastra CloudTool Rendering + Reasoning Chain
Sequential tool calls rendered with per-tool components plus a reasoning chain rendered inline via a custom reasoningMessage slot
MastraMastra CloudGen UI Interrupt (Frontend Tool)
In-chat time-picker via useFrontendTool with an async handler that blocks until the user picks a slot — Strategy B adaptation of the LangGraph interrupt primitive
MastraMastra CloudHeadless Interrupt (Frontend Tool)
Time-picker popup rendered outside the chat in the app surface — headless variant of the Strategy B interrupt adaptation
CrewAI (Crews)CrewAI EnterpriseCLI Start Command
Copy-paste command to clone the canonical starter
CrewAI (Crews)CrewAI EnterpriseAgentic Chat
Natural conversation with frontend tool execution
CrewAI (Crews)CrewAI EnterpriseReasoning
Visible reasoning/thinking chain alongside the final answer
CrewAI (Crews)CrewAI EnterpriseReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
CrewAI (Crews)CrewAI EnterprisePre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
CrewAI (Crews)CrewAI EnterprisePre-Built: Popup
Floating popup chat via <CopilotPopup />
CrewAI (Crews)CrewAI EnterpriseChat Customization (Slots)
Customize CopilotChat via its slot system
CrewAI (Crews)CrewAI EnterpriseChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
CrewAI (Crews)CrewAI EnterpriseHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
CrewAI (Crews)CrewAI EnterpriseHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
CrewAI (Crews)CrewAI EnterpriseIn-Chat Human in the Loop
User approves agent actions before execution
CrewAI (Crews)CrewAI EnterpriseIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
CrewAI (Crews)CrewAI EnterpriseIn-App Human in the Loop
Agent requests approval via useFrontendTool; the approval UI pops up as an app-level modal OUTSIDE the chat
CrewAI (Crews)CrewAI EnterpriseIn-Chat Human in the Loop (original)
Original CrewAI HITL demo — kept for backwards-compat
CrewAI (Crews)CrewAI EnterpriseTool Rendering
Custom per-tool renderers plus a wildcard catch-all for every other tool
CrewAI (Crews)CrewAI EnterpriseTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
CrewAI (Crews)CrewAI EnterpriseTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
CrewAI (Crews)CrewAI EnterpriseTool Rendering + Reasoning Chain
Sequential tool calls with reasoning tokens side-by-side
CrewAI (Crews)CrewAI EnterpriseAgentic Generative UI
Long-running agent tasks with generated UI
CrewAI (Crews)CrewAI EnterpriseTool-Based Generative UI
Agent uses tools to trigger UI generation
CrewAI (Crews)CrewAI EnterpriseFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
CrewAI (Crews)CrewAI EnterpriseFrontend Tools (Async)
useFrontendTool with an async handler
CrewAI (Crews)CrewAI EnterpriseState Streaming
Per-token state delta streaming from agent to UI
CrewAI (Crews)CrewAI EnterpriseShared State (Read + Write)
Bidirectional shared state - UI writes preferences, agent writes notes
CrewAI (Crews)CrewAI EnterpriseSub-Agents
Supervisor delegates to research / writing / critique CrewAI sub-crews with a live delegation log
CrewAI (Crews)CrewAI EnterpriseReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
CrewAI (Crews)CrewAI EnterpriseFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
CrewAI (Crews)CrewAI EnterpriseOpen-Ended Gen UI (Advanced)
Agent-authored UI that invokes host sandbox functions from inside the iframe
CrewAI (Crews)CrewAI EnterpriseAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent
CrewAI (Crews)CrewAI EnterpriseAuthentication
Runtime-level bearer-token auth via the V2 onRequest hook
CrewAI (Crews)CrewAI EnterpriseVoice Input
Speech-to-text transcription via @copilotkit/voice mounted on a per-demo runtime
CrewAI (Crews)CrewAI EnterpriseMultimodal Attachments
Image and PDF uploads routed through the AG-UI attachment pipeline
CrewAI (Crews)CrewAI EnterpriseDeclarative Generative UI (A2UI - Dynamic Schema)
Branded A2UI components generated by a secondary LLM against a custom catalog
CrewAI (Crews)CrewAI EnterpriseA2UI Fixed-Schema
Pre-authored component tree; the agent streams only data into the data model
CrewAI (Crews)CrewAI EnterpriseBYOC (Hashbrown)
Streaming structured output via @hashbrownai/react against a custom catalog
CrewAI (Crews)CrewAI EnterpriseBYOC (json-render)
Flat {root, elements} spec rendered by @json-render/react against a Zod catalog
CrewAI (Crews)CrewAI EnterpriseBeautiful Chat
Flagship showcase - polished brand-styled chat with A2UI, Open Generative UI, and shared-state todos
CrewAI (Crews)CrewAI EnterpriseMCP Apps
MCP server-driven UI via activity renderers
CrewAI (Crews)CrewAI EnterpriseGen UI Interrupt (Strategy B)
Interrupt-style scheduling via useFrontendTool with an async Promise — time-picker card rendered inline in chat
CrewAI (Crews)CrewAI EnterpriseHeadless Interrupt (Strategy B)
Headless interrupt via useFrontendTool — time-picker popup appears in the app surface outside the chat
Agent Frameworks
Google ADKGoogle AI StudioCLI Start Command
Copy-paste command to clone the canonical starter
Google ADKGoogle AI StudioPre-Built: CopilotChat
Vanilla CopilotChat with three starter-prompt suggestions — the minimum-viable surface for free-form chat
Google ADKGoogle AI StudioChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
Google ADKGoogle AI StudioTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
Google ADKGoogle AI StudioTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
Google ADKGoogle AI StudioFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
Google ADKGoogle AI StudioFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation and uses the returned result
Google ADKGoogle AI StudioHuman In the Loop: In-chat
Time-slot picker rendered inline in the chat via `useHumanInTheLoop` — the agent requests user input, the frontend renders a TimePickerCard, and the user's choice resumes the agent.
Google ADKGoogle AI StudioMCP Apps
Runtime mcpApps config wires an Excalidraw MCP server; the built-in MCPAppsActivityRenderer renders the sandboxed iframe inline in the chat with no app-side renderer
Google ADKGoogle AI StudioIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via a frontend tool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
Google ADKGoogle AI StudioTool Rendering
Custom per-tool renderers (WeatherCard, FlightListCard) plus a wildcard catch-all for every other tool
Google ADKGoogle AI StudioTool-Based Generative UI
Agent uses tools to trigger UI generation
Google ADKGoogle AI StudioAgentic Generative UI
Canonical agent-state-driven Gen UI — the agent plans a live step list and emits it via state["steps"]; the frontend renders it via useAgent
Google ADKGoogle AI StudioShared State: Read-only
Recipe editor publishes form state via agent.setState; the agent reads the recipe context but does not mutate it (no backend tool — neutral default agent).
Google ADKGoogle AI StudioShared State: Read + Write
Bidirectional shared state — UI writes preferences via agent.setState; agent writes notes via a Command-returning tool
Google ADKGoogle AI StudioState Streaming
Per-token state delta streaming from agent to UI via PredictStateMapping
Google ADKGoogle AI StudioSub-Agents
Multiple agents with visible task delegation
Google ADKGoogle AI StudioPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
Google ADKGoogle AI StudioPre-Built: Popup
Floating popup chat via <CopilotPopup />
Google ADKGoogle AI StudioChat Customization (Slots)
Customize CopilotChat via its slot system
Google ADKGoogle AI StudioHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
Google ADKGoogle AI StudioHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
Google ADKGoogle AI StudioReasoning: Custom
Visible reasoning/thinking chain alongside the final answer (Gemini 2.5 thinking mode)
Google ADKGoogle AI StudioReasoning: Default
Built-in CopilotChatReasoningMessage rendering with no slot override.
Google ADKGoogle AI StudioDeclarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; the runtime injects the render_a2ui tool automatically
Google ADKGoogle AI StudioDeclarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
Google ADKGoogle AI StudioReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
Google ADKGoogle AI StudioTool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
Google ADKGoogle AI StudioDeclarative UI: json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
Google ADKGoogle AI StudioBeautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
Google ADKGoogle AI StudioMultimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a Gemini multimodal agent
Google ADKGoogle AI StudioAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
Google ADKGoogle AI StudioDeclarative UI: Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
Google ADKGoogle AI StudioFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
Google ADKGoogle AI StudioOpen-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
Google ADKGoogle AI StudioVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
Google ADKGoogle AI StudioAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
PydanticAICLI Start Command
Copy-paste command to clone the canonical starter
PydanticAIAgentic Chat
Natural conversation with frontend tool execution
PydanticAIPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
PydanticAIPre-Built: Popup
Floating popup chat via <CopilotPopup />
PydanticAIChat Customization (Slots)
Customize CopilotChat via its slot system
PydanticAIChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
PydanticAIHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
PydanticAIFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
PydanticAIFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
PydanticAIIn-Chat Human in the Loop
User approves agent actions before execution
PydanticAIIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
PydanticAIIn-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
PydanticAIIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
PydanticAITool Rendering
Backend agent tools rendered as UI components
PydanticAITool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
PydanticAITool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
PydanticAIAgentic Generative UI
Long-running agent tasks with generated UI
PydanticAITool-Based Generative UI
Frontend-registered chart components rendered by agent tool calls
PydanticAIState Streaming
Per-token state delta streaming from agent to UI
PydanticAIShared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back
PydanticAISub-Agents
Multiple agents with visible task delegation
PydanticAIReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
PydanticAIOpen Generative UI (Minimal)
Agent streams HTML/CSS into a sandboxed iframe via the runtime's openGenerativeUI middleware
PydanticAIOpen Generative UI (Advanced)
Sandboxed UI invokes host-registered sandbox functions via Websandbox.connection.remote
PydanticAIDeclarative Generative UI (A2UI Dynamic)
Agent generates dynamic A2UI surfaces from a frontend-registered component catalog
PydanticAIA2UI — Fixed Schema
Backend-authored JSON schema + frontend catalog renderers; agent streams only data
PydanticAIHeadless Chat (Complete)
Full custom chat built from scratch on useAgent — no CopilotChat primitives
PydanticAIBeautiful Chat
Flagship showcase combining A2UI (fixed + dynamic), Open Generative UI, shared todos state, and HITL in a single polished cell
PydanticAIBYOC: json-render
BYOC (bring-your-own-components) demo rendering a streaming JSON spec via @json-render/react against a Zod-validated catalog
PydanticAIBYOC: Hashbrown
BYOC demo using @hashbrownai/react's useJsonParser + useUiKit for progressive streaming-JSON UI rendering
PydanticAIMultimodal Attachments
Image + PDF attachments via CopilotChat — images flow natively to GPT-4o vision; PDFs flatten to text via pypdf
PydanticAIVoice Input
Audio transcription via OpenAI Whisper — mic + bundled sample round-trip populate the chat composer
PydanticAIAgent Config Object
Forwarded provider properties (tone / expertise / responseLength) reshape the agent's system prompt per turn
PydanticAIAuthentication
Bearer-token gated runtime route via the V2 onRequest hook — sign out exercises the 401 path without white-screening
PydanticAIAgentic Chat (Reasoning)
Reasoning chain rendered via a custom amber slot — backed by a gpt-5 reasoning model whose Responses-API reasoning items the AG-UI bridge translates to REASONING events
PydanticAIReasoning (Default Render)
Built-in CopilotChatReasoningMessage card rendering REASONING events emitted by a gpt-5 reasoning agent — zero-config frontend
PydanticAITool Rendering (Reasoning Chain)
Reasoning + sequential tool calls (get_weather, search_flights, get_stock_price, roll_dice) in a single cell — gpt-5 reasoning model emits REASONING events between each tool call
PydanticAIMCP Apps
No-tools agent + runtime mcpApps middleware pointing at the public Excalidraw MCP server — the built-in MCPAppsActivityRenderer paints sandboxed iframes inline
PydanticAIIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
PydanticAIHeadless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop
Claude Agent SDK (Python)CLI Start Command
Copy-paste command to clone the canonical starter
Claude Agent SDK (Python)Agentic Chat
Natural conversation with frontend tool execution
Claude Agent SDK (Python)Tool Rendering
Backend agent tools rendered as UI components
Claude Agent SDK (Python)Agentic Generative UI
Long-running agent tasks with generated UI
Claude Agent SDK (Python)State Streaming
Per-token state delta streaming from agent to UI
Claude Agent SDK (Python)Shared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back
Claude Agent SDK (Python)Sub-Agents
Multiple agents with visible task delegation
Claude Agent SDK (Python)Pre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
Claude Agent SDK (Python)Pre-Built: Popup
Floating popup chat via <CopilotPopup />
Claude Agent SDK (Python)Chat Customization (Slots)
Customize CopilotChat via its slot system
Claude Agent SDK (Python)Chat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
Claude Agent SDK (Python)Headless Chat (Simple)
Minimal custom chat surface built on useAgent
Claude Agent SDK (Python)Headless Chat (Complete)
Full chat implementation built from scratch on useAgent
Claude Agent SDK (Python)Frontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
Claude Agent SDK (Python)Frontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
Claude Agent SDK (Python)In-Chat Human in the Loop
User approves agent actions before execution
Claude Agent SDK (Python)In-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
Claude Agent SDK (Python)Readonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
Claude Agent SDK (Python)Tool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
Claude Agent SDK (Python)Tool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
Claude Agent SDK (Python)Fully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
Claude Agent SDK (Python)Open-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
Claude Agent SDK (Python)BYOC: json-render
Agent emits a JSON spec; `@json-render/react` renders it against a Zod catalog of MetricCard, BarChart, PieChart
Claude Agent SDK (Python)BYOC: Hashbrown
Streaming structured output via `@hashbrownai/react` — agent emits a catalog-constrained JSON envelope that renders progressively
Claude Agent SDK (Python)Multimodal Attachments
Image + PDF attachments routed through Claude's native vision + pypdf text flattening
Claude Agent SDK (Python)Voice Input
Audio transcription wired on a per-demo runtime with a guarded OpenAI Whisper service that maps missing keys to 401
Claude Agent SDK (Python)Agent Config Object
Forwarded props (tone / expertise / responseLength) flow from the provider into the agent's system prompt via `forwardedProps.config.configurable.properties`
Claude Agent SDK (Python)Agentic Chat (Reasoning)
Visible reasoning / thinking chain rendered via a custom messageView slot
Claude Agent SDK (Python)Reasoning (Default Render)
Reasoning rendered via CopilotKit's built-in collapsible card with zero custom config
Claude Agent SDK (Python)Tool Rendering (Reasoning Chain)
Sequential tool calls combined with visible reasoning steps
Claude Agent SDK (Python)Authentication
Framework-native request auth via the V2 runtime's `onRequest` hook; missing bearer token returns 401
Claude Agent SDK (Python)MCP Apps
MCP server-driven UI via activity renderers — the agent calls a tool on a remote MCP server (Excalidraw) and the runtime middleware renders the returned UI resource as a sandboxed iframe inline in the chat
Claude Agent SDK (Python)Beautiful Chat
Polished brand-themed chat surface over the Claude Agent SDK backend with seeded suggestion pills and a glanceable decorative side panel — pure frontend dressing, no extra runtime config
Claude Agent SDK (Python)Tool-Based Generative UI
Agent uses tools to trigger UI generation (haiku generator via useFrontendTool + useRenderTool)
Claude Agent SDK (Python)In-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
Claude Agent SDK (Python)In-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
Claude Agent SDK (Python)In-Chat HITL (useInterrupt — low-level primitive)
Interactive time-picker component rendered inline in the chat via useFrontendTool with an async handler — the Promise resolves only when the user picks a slot or cancels
Claude Agent SDK (Python)Headless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop; the time-picker popup appears in the app surface outside the chat
Claude Agent SDK (Python)Declarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton/PieChart/BarChart) wired via a2ui.catalog on the provider; backend agent owns the `generate_a2ui` tool with injectA2UITool false
Claude Agent SDK (Python)Declarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema; the backend ships flight_schema.json and only streams data via display_flight
Claude Agent SDK (TypeScript)CLI Start Command
Copy-paste command to clone the canonical starter
Claude Agent SDK (TypeScript)Agentic Chat
Natural conversation with frontend tool execution
Claude Agent SDK (TypeScript)Pre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
Claude Agent SDK (TypeScript)Pre-Built: Popup
Floating popup chat via <CopilotPopup />
Claude Agent SDK (TypeScript)Chat Customization (Slots)
Customize CopilotChat via its slot system
Claude Agent SDK (TypeScript)Chat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
Claude Agent SDK (TypeScript)Headless Chat (Simple)
Minimal custom chat surface built on useAgent
Claude Agent SDK (TypeScript)Frontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
Claude Agent SDK (TypeScript)Frontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
Claude Agent SDK (TypeScript)In-Chat Human in the Loop
User approves agent actions before execution
Claude Agent SDK (TypeScript)In-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
Claude Agent SDK (TypeScript)Tool Rendering
Backend agent tools rendered as UI components
Claude Agent SDK (TypeScript)Declarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog wired via a2ui.catalog on the provider; runtime injects the render_a2ui tool automatically
Claude Agent SDK (TypeScript)Agentic Generative UI
Long-running agent tasks with generated UI
Claude Agent SDK (TypeScript)Tool-Based Generative UI
Agent uses frontend tools to trigger UI generation — useFrontendTool with a render function paints structured haiku output as a styled card alongside the chat
Claude Agent SDK (TypeScript)State Streaming
Per-token state delta streaming from agent to UI
Claude Agent SDK (TypeScript)Shared State (Read + Write)
Bidirectional shared state — UI writes preferences via agent.setState, agent reads them every turn and writes notes back via a set_notes tool
Claude Agent SDK (TypeScript)Sub-Agents
Supervisor delegates to research / writing / critique sub-agents (each a single secondary Anthropic Messages call); every delegation appears in a live log via shared state
Claude Agent SDK (TypeScript)Readonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
Claude Agent SDK (TypeScript)Fully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
Claude Agent SDK (TypeScript)Open-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
Claude Agent SDK (TypeScript)BYOC: json-render
Bring-your-own-catalog generative UI via @json-render/react
Claude Agent SDK (TypeScript)BYOC: hashbrown
Bring-your-own-catalog streaming UI via @hashbrownai/react useJsonParser
Claude Agent SDK (TypeScript)MCP Apps
MCP server-driven UI via activity renderers — runtime auto-applies the MCP Apps middleware and the built-in MCPAppsActivityRenderer paints sandboxed iframes inline in the chat
Claude Agent SDK (TypeScript)Multimodal Attachments
Image and PDF attachments forwarded to Claude's vision-capable model
Claude Agent SDK (TypeScript)Voice Input
Mic + sample audio transcription via a guarded OpenAI Whisper service
Claude Agent SDK (TypeScript)Agent Config Object
forwardedProps route provider-configured tone/expertise/length into the Claude system prompt per turn
Claude Agent SDK (TypeScript)Authentication
Bearer-token gate via the V2 runtime onRequest hook with typed 401 surface
Claude Agent SDK (TypeScript)Tool Rendering: Default Catch-all
Zero-config wildcard renderer via useDefaultRenderTool — built-in tool-call card paints every tool
Claude Agent SDK (TypeScript)Tool Rendering: Custom Catch-all
Single branded wildcard renderer registered via useDefaultRenderTool({ render })
Claude Agent SDK (TypeScript)Agentic Chat (Reasoning)
Custom reasoningMessage slot renders Claude extended-thinking deltas as AG-UI REASONING_MESSAGE_* events
Claude Agent SDK (TypeScript)Reasoning (Default Render)
Built-in CopilotChatReasoningMessage card paints Claude extended-thinking output — zero custom slots
Claude Agent SDK (TypeScript)Tool Rendering: Reasoning Chain
Reasoning slot + per-tool renderers + wildcard catch-all in one cell
Claude Agent SDK (TypeScript)Beautiful Chat
Polished brand-themed chat surface over the shared Claude SDK agent with seeded suggestion pills
Claude Agent SDK (TypeScript)In-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
Claude Agent SDK (TypeScript)In-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
Claude Agent SDK (TypeScript)In-Chat HITL (useInterrupt — low-level primitive)
Interactive time-picker component rendered inline in the chat via useFrontendTool with an async handler — the Promise resolves only when the user picks a slot or cancels
Claude Agent SDK (TypeScript)Headless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop; the time-picker popup appears in the app surface outside the chat
Claude Agent SDK (TypeScript)Headless Chat (Complete)
Full custom chat surface built on `useAgent` — exercises the full rendering stack (text, per-tool renderers, frontend components) with backend get_weather / get_stock_price tools and a frontend highlight_note tool
Claude Agent SDK (TypeScript)A2UI (Fixed Schema)
Fixed-schema A2UI — backend display_flight tool emits an a2ui_operations container that ships a JSON component tree plus per-call data; frontend catalog binds component names to React renderers
Enterprise
MS Agent Framework (Python)CLI Start Command
Copy-paste command to clone the canonical starter
MS Agent Framework (Python)Agentic Chat
Natural conversation with frontend tool execution
MS Agent Framework (Python)Chat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
MS Agent Framework (Python)Tool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
MS Agent Framework (Python)Tool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
MS Agent Framework (Python)Frontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
MS Agent Framework (Python)Frontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
MS Agent Framework (Python)In-Chat Human in the Loop
User approves agent actions before execution
MS Agent Framework (Python)In-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
MS Agent Framework (Python)In-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
MS Agent Framework (Python)In-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
MS Agent Framework (Python)Tool-Based Generative UI
Agent uses tools to trigger UI generation
MS Agent Framework (Python)Tool Rendering
Backend agent tools rendered as UI components
MS Agent Framework (Python)Agentic Generative UI
Long-running agent tasks with generated UI
MS Agent Framework (Python)Shared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back via state_update
MS Agent Framework (Python)Sub-Agents
Supervisor delegates to research / writing / critique sub-agents with a live delegation log
MS Agent Framework (Python)State Streaming
Per-token state delta streaming from agent to UI
MS Agent Framework (Python)Pre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
MS Agent Framework (Python)Pre-Built: Popup
Floating popup chat via <CopilotPopup />
MS Agent Framework (Python)Chat Customization (Slots)
Customize CopilotChat via its slot system
MS Agent Framework (Python)Headless Chat (Simple)
Minimal custom chat surface built on useAgent
MS Agent Framework (Python)Headless Chat (Complete)
Full chat implementation built from scratch on useAgent
MS Agent Framework (Python)Reasoning
Visible reasoning/thinking chain alongside the final answer
MS Agent Framework (Python)In-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
MS Agent Framework (Python)Declarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; runtime injects the render_a2ui tool automatically
MS Agent Framework (Python)Declarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
MS Agent Framework (Python)MCP Apps
MCP server-driven UI via activity renderers
MS Agent Framework (Python)Headless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop
MS Agent Framework (Python)Readonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
MS Agent Framework (Python)Reasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
MS Agent Framework (Python)Tool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
MS Agent Framework (Python)BYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
MS Agent Framework (Python)Beautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
MS Agent Framework (Python)Multimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
MS Agent Framework (Python)Authentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
MS Agent Framework (Python)BYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
MS Agent Framework (Python)Fully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
MS Agent Framework (Python)Open-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
MS Agent Framework (Python)Voice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
MS Agent Framework (Python)Agent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
MS Agent Framework (.NET)CLI Start Command
Copy-paste command to clone the canonical starter
MS Agent Framework (.NET)Agentic Chat
Natural conversation with frontend tool execution
MS Agent Framework (.NET)Chat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
MS Agent Framework (.NET)Tool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
MS Agent Framework (.NET)Tool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
MS Agent Framework (.NET)Frontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
MS Agent Framework (.NET)Frontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
MS Agent Framework (.NET)In-Chat Human in the Loop (Original)
User approves agent actions before execution
MS Agent Framework (.NET)In-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
MS Agent Framework (.NET)In-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
MS Agent Framework (.NET)In-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
MS Agent Framework (.NET)Tool-Based Generative UI
Agent uses tools to trigger UI generation
MS Agent Framework (.NET)Tool Rendering
Backend agent tools rendered as UI components
MS Agent Framework (.NET)Agentic Generative UI
Long-running agent tasks with generated UI
MS Agent Framework (.NET)State Streaming
Per-token state delta streaming from agent to UI
MS Agent Framework (.NET)Shared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back via a set_notes tool
MS Agent Framework (.NET)Sub-Agents
Supervisor delegates to research / writing / critique sub-agents with a live delegation log
MS Agent Framework (.NET)Pre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
MS Agent Framework (.NET)Pre-Built: Popup
Floating popup chat via <CopilotPopup />
MS Agent Framework (.NET)Chat Customization (Slots)
Customize CopilotChat via its slot system
MS Agent Framework (.NET)Headless Chat (Simple)
Minimal custom chat surface built on useAgent
MS Agent Framework (.NET)Headless Chat (Complete)
Full chat implementation built from scratch on useAgent
MS Agent Framework (.NET)Reasoning
Visible reasoning/thinking chain alongside the final answer
MS Agent Framework (.NET)In-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
MS Agent Framework (.NET)Declarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; runtime injects the render_a2ui tool automatically
MS Agent Framework (.NET)Declarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
MS Agent Framework (.NET)MCP Apps
MCP server-driven UI via activity renderers
MS Agent Framework (.NET)Headless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop
MS Agent Framework (.NET)Readonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
MS Agent Framework (.NET)Reasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
MS Agent Framework (.NET)Tool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
MS Agent Framework (.NET)BYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
MS Agent Framework (.NET)Beautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
MS Agent Framework (.NET)Multimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
MS Agent Framework (.NET)Authentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
MS Agent Framework (.NET)BYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
MS Agent Framework (.NET)Fully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
MS Agent Framework (.NET)Open-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
MS Agent Framework (.NET)Voice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
MS Agent Framework (.NET)Agent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
Spring AIAgentic Chat
Natural conversation with frontend tool execution
Spring AIChat Customization (CSS)
Default CopilotChat re-themed via scoped CSS overrides
Spring AIFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
Spring AIFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation and uses the returned result
Spring AIAgentic Generative UI
Long-running agent tasks with generated UI
Spring AIPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
Spring AIPre-Built: Popup
Floating popup chat via <CopilotPopup />
Spring AIChat Customization (Slots)
Customize CopilotChat via its slot system
Spring AIHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
Spring AIReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
Spring AIAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
Spring AIOpen Generative UI (Minimal)
Agent streams HTML + CSS into a sandboxed iframe via the built-in OpenGenerativeUIActivityRenderer
Spring AIOpen Generative UI (Advanced)
Sandboxed UI can call host-side functions (evaluateExpression, notifyHost) via the provider's sandboxFunctions registration
Spring AIVoice
CopilotChat mic button via a runtime mounted with TranscriptionServiceOpenAI; transcribed text is forwarded to the Spring AI backend
Spring AIAgent Config Object
Frontend forwards typed config (tone, expertise, responseLength) to Spring, which rebuilds the system prompt per request
Spring AIA2UI Fixed Schema
Dedicated Spring tool emits a fixed flight-card component tree plus a data model; frontend catalog renders each component
Spring AIHeadless Chat (Complete)
Full chat built from scratch on useAgent plus low-level render hooks (useRenderToolCall, useRenderActivityMessage, useRenderCustomMessages)
Spring AIBeautiful Chat
Polished brand-themed chat surface over the Spring AI agent with seeded suggestion pills
Spring AIAgentic Chat (Reasoning)
Custom reasoningMessage slot renders the agent's thinking chain as a tagged amber banner (renders when the adapter emits REASONING_MESSAGE_* events)
Spring AIReasoning (Default Render)
Zero-config reasoning — CopilotChatReasoningMessage renders the built-in collapsible card when REASONING_MESSAGE_* events are emitted
Spring AITool Rendering (Reasoning Chain)
Per-tool renderers (WeatherCard, FlightListCard) plus custom reasoningMessage slot composed in a single chat
Spring AIDeclarative Generative UI
A2UI dynamic-schema against a branded catalog; backend owns the `generate_a2ui` tool
Spring AIShared State (Read + Write)
Bidirectional shared state — UI writes preferences into agent state; the agent's set_notes tool mutates state.notes and emits a STATE_SNAPSHOT back to the UI
Spring AIShared State Streaming
Per-token state streaming — the agent's write_document tool streams content into state.document via STATE_SNAPSHOT events; the frontend renders the growing document live
Spring AISub-Agents
Supervisor delegates to research / writing / critique sub-agents (each its own ChatClient call). Every delegation appends a Delegation entry to state.delegations and emits a STATE_SNAPSHOT for a live UI log
Spring AIMultimodal Attachments
CopilotChat AttachmentsConfig for images + PDFs; sample files inject through the same pipeline as the paperclip upload
Spring AIIn-Chat Human in the Loop
User approves agent actions before execution via text input inside the chat
Spring AIIn-App Human in the Loop
Agent requests approval via useFrontendTool; the approval UI pops up as an app-level modal OUTSIDE the chat
Spring AIHuman in the Loop (Steps)
Multi-step HITL flow where the agent pauses for user confirmation at each step
Spring AITool Rendering
Backend agent tools rendered as UI components
Spring AICLI Start Command
Copy-paste command to clone the canonical Spring AI starter
Spring AITool-Based Generative UI
Frontend registers per-tool component renderers via useComponent; the Spring agent calls render tools that paint chart components inline in the chat
Spring AITool Rendering (Default Catch-all)
Out-of-the-box tool rendering — the Spring backend defines tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
Spring AITool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every Spring tool call
Spring AIIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow against the Spring AI agent
Spring AIMCP Apps
MCP server-driven UI via the runtime mcpApps middleware; sandboxed iframe rendered via the built-in MCPAppsActivityRenderer
Spring AIBYOC Hashbrown
Streaming hashbrown UI kit (MetricCard + PieChart + BarChart + DealCard) rendered by @hashbrownai/react against the Spring AI ChatClient stream
Spring AIGenerative UI (Interrupt)
Time-picker card rendered inline via useFrontendTool with an async handler that blocks until the user picks a slot — Strategy B adaptation of LangGraph's interrupt() primitive
Spring AIHeadless Interrupt
Time-picker popup rendered in the app surface (outside the chat) via useFrontendTool with an async handler — headless variant of the interrupt-adapted pattern
Spring AIBYOC JSON Render
Zero-tool agent emits a flat-spec JSON object rendered by @json-render/react against a Zod-validated catalog of MetricCard, BarChart, and PieChart components
Emerging
LangGraph (FastAPI)LangGraph PlatformCLI Start Command
Copy-paste command to clone the canonical starter
LangGraph (FastAPI)LangGraph PlatformAgentic Chat
Natural conversation with frontend tool execution
LangGraph (FastAPI)LangGraph PlatformTool Rendering
Backend agent tools rendered as UI components
LangGraph (FastAPI)LangGraph PlatformIn-Chat Human in the Loop (Original)
User approves agent actions before execution
LangGraph (FastAPI)LangGraph PlatformAgentic Generative UI
Long-running agent tasks with generated UI
LangGraph (FastAPI)LangGraph PlatformTool-Based Generative UI
Agent uses tools to trigger UI generation
LangGraph (FastAPI)LangGraph PlatformState Streaming
Per-token state delta streaming from agent to UI
LangGraph (FastAPI)LangGraph PlatformShared State (Read + Write)
Bidirectional shared state — UI writes preferences, agent writes notes
LangGraph (FastAPI)LangGraph PlatformSub-Agents
Multiple agents with visible task delegation
LangGraph (FastAPI)LangGraph PlatformPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
LangGraph (FastAPI)LangGraph PlatformPre-Built: Popup
Floating popup chat via <CopilotPopup />
LangGraph (FastAPI)LangGraph PlatformChat Customization (Slots)
Customize CopilotChat via its slot system
LangGraph (FastAPI)LangGraph PlatformChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
LangGraph (FastAPI)LangGraph PlatformBeautiful Chat
Canonical polished starter chat — brand fonts, theme tokens, suggestion pills
LangGraph (FastAPI)LangGraph PlatformHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
LangGraph (FastAPI)LangGraph PlatformHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
LangGraph (FastAPI)LangGraph PlatformAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
LangGraph (FastAPI)LangGraph PlatformMultimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
LangGraph (FastAPI)LangGraph PlatformAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
LangGraph (FastAPI)LangGraph PlatformReasoning
Visible reasoning/thinking chain alongside the final answer
LangGraph (FastAPI)LangGraph PlatformReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
LangGraph (FastAPI)LangGraph PlatformTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
LangGraph (FastAPI)LangGraph PlatformTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
LangGraph (FastAPI)LangGraph PlatformTool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
LangGraph (FastAPI)LangGraph PlatformIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the LangGraph interrupt lifecycle
LangGraph (FastAPI)LangGraph PlatformHeadless Interrupt (testing)
Resolve langgraph interrupts from a plain button grid — no chat, no useInterrupt render prop
LangGraph (FastAPI)LangGraph PlatformDeclarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton) wired via a2ui.catalog on the provider; runtime injects the render_a2ui tool automatically
LangGraph (FastAPI)LangGraph PlatformDeclarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
LangGraph (FastAPI)LangGraph PlatformMCP Apps
MCP server-driven UI via activity renderers
LangGraph (FastAPI)LangGraph PlatformFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
LangGraph (FastAPI)LangGraph PlatformFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
LangGraph (FastAPI)LangGraph PlatformIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat, and a completion callback resolves the pending tool Promise with the user's decision
LangGraph (FastAPI)LangGraph PlatformIn-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
LangGraph (FastAPI)LangGraph PlatformIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
LangGraph (FastAPI)LangGraph PlatformReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
LangGraph (FastAPI)LangGraph PlatformVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
LangGraph (FastAPI)LangGraph PlatformBYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
LangGraph (FastAPI)LangGraph PlatformBYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
LangGraph (FastAPI)LangGraph PlatformFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library
LangGraph (FastAPI)LangGraph PlatformOpen-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
AgnoAgent OSCLI Start Command
Copy-paste command to clone the canonical starter
AgnoAgent OSAgentic Chat
Natural conversation with frontend tool execution
AgnoAgent OSTool Rendering
Backend agent tools rendered as UI components
AgnoAgent OSAgentic Generative UI
Long-running agent tasks with generated UI
AgnoAgent OSState Streaming
Per-token state delta streaming from agent to UI
AgnoAgent OSShared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back
AgnoAgent OSSub-Agents
Multiple agents with visible task delegation
AgnoAgent OSPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
AgnoAgent OSPre-Built: Popup
Floating popup chat via <CopilotPopup />
AgnoAgent OSChat Customization (Slots)
Customize CopilotChat via its slot system
AgnoAgent OSChat Customization (CSS)
Default CopilotChat re-themed via CopilotKit CSS variables
AgnoAgent OSHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
AgnoAgent OSIn-Chat Human in the Loop
User approves agent actions before execution
AgnoAgent OSIn-Chat Human in the Loop (Time Picker)
Agent proposes a meeting time; user confirms or adjusts via an in-chat card before the agent proceeds
AgnoAgent OSIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat
AgnoAgent OSFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
AgnoAgent OSReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
AgnoAgent OSTool Rendering (Default Catch-all)
Out-of-the-box default tool-call card via useDefaultRenderTool()
AgnoAgent OSTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool
AgnoAgent OSFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side notes-DB query and uses the returned result
AgnoAgent OSAgentic Chat (Reasoning)
Visible reasoning chain via a custom ReasoningBlock slot — Agno reasoning agent emits AG-UI REASONING_MESSAGE_* events
AgnoAgent OSReasoning (Default Render)
Zero-config reasoning rendering — CopilotKit's built-in CopilotChatReasoningMessage handles the AGUI reasoning events
AgnoAgent OSTool Rendering (Reasoning Chain)
Reasoning card + sequential tool-call renderers (per-tool + catch-all) on the same reasoning-enabled Agno agent
AgnoAgent OSHeadless Chat (Complete)
Full chat built from scratch on useAgent — custom bubbles, input bar, and manual generative-UI composition via useRenderToolCall
AgnoAgent OSGen UI Interrupt (Frontend Tool + async Promise)
In-chat time-picker card via useFrontendTool with an async handler that blocks until the user picks a slot — Agno adaptation of the LangGraph interrupt() primitive
AgnoAgent OSHeadless Interrupt (Frontend Tool + async Promise)
Time-picker popup rendered outside the chat in the app surface via useFrontendTool with an async handler — Agno adaptation of the LangGraph headless interrupt pattern
AgnoAgent OSAuthentication
Bearer-token gate via the V2 runtime onRequest hook — dedicated /api/copilotkit-auth route rejects unauthenticated requests
AgnoAgent OSTool-Based Generative UI
Agent uses tools to trigger UI generation — frontend renders a haiku card from a generate_haiku tool call
AgnoAgent OSIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow — same surface as hitl-in-chat, branded for the booking scenario
AgnoAgent OSMCP Apps
MCP-provided tools rendered inline as sandboxed UI via the runtime's mcpApps middleware
AgnoAgent OSOpen Generative UI
Agent-authored HTML + CSS streamed into a sandboxed iframe via the runtime's openGenerativeUI middleware
AgnoAgent OSOpen Generative UI (Advanced)
Sandboxed iframe with host-side sandbox functions the agent's UI can call back into via Websandbox.connection.remote
AgnoAgent OSAgent Config Object
Frontend forwards a typed config object (tone, expertise, response length) the Agno agent reads per turn to compose its system prompt
AgnoAgent OSVoice Input
Microphone + sample-audio button → /transcribe endpoint → text injected into the chat composer
AgnoAgent OSMultimodal Attachments
Image + PDF attachments forwarded to a vision-capable Agno agent via CopilotChat's attachments config
AgnoAgent OSBYOC: Hashbrown
Streaming structured output rendered progressively via @hashbrownai/react useUiKit + useJsonParser
AgnoAgent OSBYOC: JSON Render
Agent emits a json-render spec; the page renders it against a Zod-validated catalog (MetricCard, BarChart, PieChart)
AgnoAgent OSBeautiful Chat
Polished landing-style chat shell over the shared Agno agent — gradient background, suggestion pills, and an example dashboard surface
AgnoAgent OSDeclarative Generative UI (A2UI Dynamic)
Agent dynamically composes UI per turn from a registered frontend catalog — secondary LLM call inside the backend `generate_a2ui` tool emits an a2ui_operations container that the A2UI middleware forwards to the catalog renderer
AgnoAgent OSA2UI Fixed Schema
Schema is authored ahead of time as JSON; the agent only streams data into the data model. `display_flight` emits an a2ui_operations container directly with no secondary LLM call
AG2Agent OSCLI Start Command
Copy-paste command to clone the canonical starter
AG2Agent OSAgentic Chat
Natural conversation with frontend tool execution
AG2Agent OSIn-Chat HITL (useHumanInTheLoop)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
AG2Agent OSIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow — same surface as hitl-in-chat with a booking-focused framing
AG2Agent OSIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat
AG2Agent OSIn-Chat Human in the Loop (Original)
User approves agent actions before execution
AG2Agent OSTool Rendering
Backend agent tools rendered as UI components
AG2Agent OSTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
AG2Agent OSTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
AG2Agent OSAgentic Generative UI
Long-running agent tasks with generated UI
AG2Agent OSTool-Based Generative UI
Agent uses tools to trigger UI generation
AG2Agent OSState Streaming
Per-token state delta streaming from agent to UI
AG2Agent OSPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
AG2Agent OSPre-Built: Popup
Floating popup chat via <CopilotPopup />
AG2Agent OSChat Customization (Slots)
Customize CopilotChat via its slot system
AG2Agent OSChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
AG2Agent OSHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
AG2Agent OSHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
AG2Agent OSReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
AG2Agent OSReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
AG2Agent OSFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
AG2Agent OSFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation and uses the returned result
AG2Agent OSShared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back via AG2 ContextVariables
AG2Agent OSSub-Agents
Supervisor delegates tasks to research, writing, and critique sub-agents with a live delegation log
AG2Agent OSGen UI Interrupt (Frontend Tool + async Promise)
In-chat time-picker card via useFrontendTool with an async handler that blocks until the user picks a slot — AG2 adaptation of the LangGraph interrupt() primitive
AG2Agent OSHeadless Interrupt (Frontend Tool + async Promise)
Time-picker popup rendered outside the chat in the app surface via useFrontendTool with an async handler — AG2 adaptation of the LangGraph headless interrupt pattern
AG2Agent OSDeclarative Generative UI (A2UI Dynamic Schema)
Agent dynamically composes UI from a registered catalog of branded components via the A2UI middleware
AG2Agent OSA2UI (Fixed Schema)
Agent streams data into a frontend-authored fixed component schema; backend ships JSON, agent fills in values
AG2Agent OSMCP Apps
Runtime mcpApps middleware injects an MCP server's tools and renders associated UI resources inline
AG2Agent OSOpen Generative UI (Minimal)
Agent streams a single generateSandboxedUi tool call; the runtime renders agent-authored HTML/CSS in a sandboxed iframe
AG2Agent OSOpen Generative UI (Advanced)
Sandboxed iframe UIs invoke host-page sandbox functions via Websandbox.connection.remote
AG2Agent OSAgentic Chat (Reasoning)
Custom reasoningMessage slot renders the agent's thinking chain visibly above each answer
AG2Agent OSTool Rendering (Reasoning Chain)
Sequential tool calls combined with a custom reasoningMessage slot — weather, flights, custom catch-all
AG2Agent OSAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
AG2Agent OSAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent's dynamic system prompt
AG2Agent OSVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
AG2Agent OSMultimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent
AG2Agent OSBYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
AG2Agent OSBYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
AG2Agent OSBeautiful Chat
Simplified flagship cell that combines A2UI Dynamic + Open Generative UI on a single dedicated runtime — agent picks branded catalog components for structured visuals and free-form sandboxed UI for everything else
LlamaIndexCLI Start Command
Copy-paste command to clone the canonical starter
LlamaIndexBeautiful Chat
Polished starter chat with brand-styled CopilotChat surface
LlamaIndexTool-Based Generative UI
Agent uses tools to trigger UI generation (bar / pie chart components)
LlamaIndexMCP Apps
MCP server-driven UI via activity renderers
LlamaIndexAgentic Chat
Natural conversation with frontend tool execution
LlamaIndexFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
LlamaIndexFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a simulated client-side notes DB query and uses the returned result
LlamaIndexIn-Chat HITL (useHumanInTheLoop)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
LlamaIndexIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
LlamaIndexIn-App Human in the Loop (Frontend Tools + async HITL)
Agent requests approval via useFrontendTool with an async handler; the approval UI pops up as an app-level modal OUTSIDE the chat
LlamaIndexIn-Chat Human in the Loop (Original)
User approves agent actions before execution
LlamaIndexTool Rendering
Backend agent tools rendered as UI components
LlamaIndexTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines tools; frontend adds zero custom renderers and uses CopilotKit's built-in default UI
LlamaIndexTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
LlamaIndexAgentic Generative UI
Long-running agent tasks with generated UI
LlamaIndexState Streaming
Per-token state delta streaming from agent to UI
LlamaIndexShared State (Read + Write)
Bidirectional shared state — UI writes preferences via agent.setState; agent writes notes via the set_notes tool. Backend reads preferences from state every turn through LlamaIndex's <state> prelude
LlamaIndexSub-Agents
Supervisor delegates to research / writing / critique sub-agents (each a stand-alone LlamaIndex FunctionAgent). Every delegation appends to a live log in shared agent state
LlamaIndexPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
LlamaIndexPre-Built: Popup
Floating popup chat via <CopilotPopup />
LlamaIndexChat Customization (Slots)
Customize CopilotChat via its slot system
LlamaIndexChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
LlamaIndexHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
LlamaIndexHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
LlamaIndexReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
LlamaIndexReasoning
Visible reasoning/thinking chain alongside the final answer via a custom reasoningMessage slot
LlamaIndexReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
LlamaIndexTool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side
LlamaIndexDeclarative Generative UI (A2UI)
Canonical A2UI BYOC — custom catalog (Card/StatusBadge/Metric/InfoRow/PrimaryButton/PieChart/BarChart) wired via a2ui.catalog on the provider; agent owns the generate_a2ui tool
LlamaIndexDeclarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema; the agent streams flight data into a pre-authored component tree
LlamaIndexBYOC json-render
Streaming hierarchical JSON UI spec rendered via @json-render/react, with a Zod-validated catalog (MetricCard + PieChart + BarChart)
LlamaIndexBYOC Hashbrown
Streaming structured output via @hashbrownai/react, rendering a sales dashboard catalog (MetricCard + PieChart + BarChart)
LlamaIndexAuthentication
Bearer-token gate via runtime onRequest hook with unauthenticated / authenticated states
LlamaIndexVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
LlamaIndexAgent Config Object
Forward a typed config object (tone / expertise / response length) from the provider to the agent
LlamaIndexMultimodal Attachments
Image and PDF uploads via CopilotChat attachments, processed by a vision-capable agent (gpt-4o)
LlamaIndexFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library inside a sandboxed iframe
LlamaIndexOpen-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
LlamaIndexIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
LlamaIndexHeadless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop
AWS StrandsAWS Bedrock AgentCoreCLI Start Command
Copy-paste command to clone the canonical starter
AWS StrandsAWS Bedrock AgentCoreAgentic Chat
Natural conversation with frontend tool execution
AWS StrandsAWS Bedrock AgentCoreChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
AWS StrandsAWS Bedrock AgentCorePre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
AWS StrandsAWS Bedrock AgentCorePre-Built: Popup
Floating popup chat via <CopilotPopup />
AWS StrandsAWS Bedrock AgentCoreChat Customization (Slots)
Customize CopilotChat via its slot system
AWS StrandsAWS Bedrock AgentCoreHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
AWS StrandsAWS Bedrock AgentCoreHeadless Chat (Complete)
Full chat implementation built from scratch on useAgent
AWS StrandsAWS Bedrock AgentCoreReasoning
Visible reasoning/thinking chain alongside the final answer
AWS StrandsAWS Bedrock AgentCoreReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot
AWS StrandsAWS Bedrock AgentCoreFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
AWS StrandsAWS Bedrock AgentCoreFrontend Tools (Async)
useFrontendTool with an async handler
AWS StrandsAWS Bedrock AgentCoreIn-Chat HITL (useHumanInTheLoop)
Inline approval/decision surface via the ergonomic useHumanInTheLoop hook
AWS StrandsAWS Bedrock AgentCoreIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
AWS StrandsAWS Bedrock AgentCoreIn-App Human in the Loop
Agent requests approval via async useFrontendTool; UI pops as an app-level modal
AWS StrandsAWS Bedrock AgentCoreIn-Chat HITL (Steps)
Agent pauses for user approval via step-based human-in-the-loop
AWS StrandsAWS Bedrock AgentCoreTool Rendering
Custom render for tool calls inline in the chat stream
AWS StrandsAWS Bedrock AgentCoreTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
AWS StrandsAWS Bedrock AgentCoreTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
AWS StrandsAWS Bedrock AgentCoreTool Rendering + Reasoning Chain
Sequential tool calls with reasoning tokens rendered side-by-side
AWS StrandsAWS Bedrock AgentCoreAgentic Generative UI
Long-running agent tasks with generated UI
AWS StrandsAWS Bedrock AgentCoreTool-Based Generative UI
Agent uses tools to trigger UI generation
AWS StrandsAWS Bedrock AgentCoreDeclarative Generative UI (A2UI)
Dynamic A2UI BYOC via a custom catalog
AWS StrandsAWS Bedrock AgentCoreDeclarative Generative UI (A2UI — Fixed Schema)
A2UI rendering against a known client-side schema
AWS StrandsAWS Bedrock AgentCoreShared State (Read + Write)
Bidirectional agent state — UI writes preferences, agent writes notes back
AWS StrandsAWS Bedrock AgentCoreSub-Agents
Multiple agents with visible task delegation
AWS StrandsAWS Bedrock AgentCoreState Streaming
Per-token state delta streaming from agent to UI
AWS StrandsAWS Bedrock AgentCoreReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
AWS StrandsAWS Bedrock AgentCoreMultimodal Attachments
Image and PDF uploads via CopilotChat attachments
AWS StrandsAWS Bedrock AgentCoreVoice Input
Speech-to-text via @copilotkit/voice with a bundled sample audio button
AWS StrandsAWS Bedrock AgentCoreAuthentication
Bearer-token gate via runtime onRequest hook
AWS StrandsAWS Bedrock AgentCoreAgent Config Object
Forward a typed config object (tone / expertise / response length) from provider to agent
AWS StrandsAWS Bedrock AgentCoreBYOC: Hashbrown
Streaming structured output rendered via @hashbrownai/react's useUiKit + useJsonParser
AWS StrandsAWS Bedrock AgentCoreBYOC: json-render
JSON spec rendered via @json-render/react against a Zod catalog
AWS StrandsAWS Bedrock AgentCoreOpen Generative UI (Minimal)
Agent-authored HTML/CSS/SVG in a sandboxed iframe, minimal variant
AWS StrandsAWS Bedrock AgentCoreOpen Generative UI (Advanced)
Sandboxed iframe that invokes host-registered sandbox functions via Websandbox.connection.remote
AWS StrandsAWS Bedrock AgentCoreMCP Apps
MCP server-driven UI via activity renderers
AWS StrandsAWS Bedrock AgentCoreBeautiful Chat
Polished landing-style chat surface with brand theming and suggestion pills
AWS StrandsAWS Bedrock AgentCoreIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
AWS StrandsAWS Bedrock AgentCoreHeadless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop
LangroidCLI Start Command
Copy-paste command to clone the canonical starter
LangroidAgentic Chat
Natural conversation with frontend tool execution
LangroidIn-Chat Human in the Loop
User approves agent actions before execution
LangroidIn-Chat HITL (useHumanInTheLoop — ergonomic API)
Interactive approval/decision surface rendered inline in the chat via the high-level `useHumanInTheLoop` hook
LangroidIn-Chat HITL (Booking)
Time-picker card rendered inline via useHumanInTheLoop for a booking flow
LangroidIn-App Human in the Loop
Agent requests approval via async useFrontendTool; UI pops as an app-level modal
LangroidTool Rendering
Custom render for tool calls inline in the chat stream
LangroidTool-Based Generative UI
Agent uses tools to trigger UI generation
LangroidAgentic Generative UI
Long-running agent tasks with generated UI
LangroidState Streaming
Per-token state delta streaming from agent to UI
LangroidShared State (Read + Write)
Bidirectional shared state — UI writes preferences via agent.setState, agent writes notes via the set_notes tool
LangroidSub-Agents
Supervisor LLM delegates to research / writing / critique sub-agents; live delegation log streams running -> completed transitions
LangroidChat Customization (CSS)
Default CopilotChat re-themed via CopilotKitCSSProperties
LangroidPre-Built: Sidebar
Docked sidebar chat via <CopilotSidebar />
LangroidPre-Built: Popup
Floating popup chat via <CopilotPopup />
LangroidChat Customization (Slots)
Customize CopilotChat via its slot system (welcomeScreen, disclaimer, assistantMessage)
LangroidHeadless Chat (Simple)
Minimal custom chat surface built on useAgent
LangroidFrontend Tools (In-App Actions)
Agent invokes client-side handlers registered with useFrontendTool
LangroidFrontend Tools (Async)
useFrontendTool with an async handler — agent awaits a client-side async operation (simulated notes DB query) and uses the returned result
LangroidReasoning
Custom reasoning slot render (placeholder until Langroid exposes reasoning events)
LangroidReasoning (Default Render)
Built-in CopilotChatReasoningMessage renders without a custom slot (placeholder until Langroid exposes reasoning events)
LangroidReadonly State (Agent Context)
Frontend provides read-only context to the agent via useAgentContext
LangroidDeclarative Generative UI
A2UI dynamic-schema rendering — agent emits operations against a client-declared catalog
LangroidAuthentication
Bearer-token gate via the V2 runtime's onRequest hook; unauth requests are rejected with 401
LangroidHeadless Chat (Complete)
A full chat surface built from scratch on useAgent — no CopilotChat, hand-rolled message list + bubbles
LangroidAgent Config Object
Typed config forwarded via CopilotKit provider properties; the agent reads tone/expertise/responseLength per turn
LangroidTool Rendering (Default Catch-all)
Out-of-the-box tool rendering — backend defines the tools; the frontend adds zero custom renderers and relies on CopilotKit's built-in default UI
LangroidTool Rendering (Custom Catch-all)
Single branded wildcard renderer via useDefaultRenderTool — the same app-designed card paints every tool call
LangroidTool Rendering + Reasoning Chain (testing)
Sequential tool calls with reasoning tokens rendered side-by-side. NOTE — Langroid does not yet emit native reasoning AG-UI events; the slot wires up only when reasoning content arrives over the wire
LangroidFully Open-Ended Generative UI
Agent generates UI from an arbitrary component library via the runtime's openGenerativeUI middleware
LangroidOpen-Ended Gen UI (Advanced: with frontend function calling)
Agent-authored UI that can invoke frontend sandbox functions from inside the iframe
LangroidVoice Input
Speech-to-text via @copilotkit/voice — mic button or sample audio populates the chat composer
LangroidMultimodal Attachments
Image + PDF uploads forwarded to a vision-capable agent (gpt-4o); PDFs are flattened to text server-side
LangroidBYOC: Hashbrown
Frontend BYOC via @hashbrownai/react — agent emits a JSON envelope that streams progressively into a sales dashboard
LangroidBYOC: json-render
Frontend BYOC via @json-render/react — agent emits a flat element-map spec rendered against a Zod-validated catalog
LangroidMCP Apps
MCP servers with UI resources rendered inline as sandboxed iframes via the runtime's mcpApps middleware
LangroidBeautiful Chat
Polished landing-style chat shell over the shared Langroid agent — gradient background, suggestion pills, and an example dashboard surface
LangroidA2UI (Fixed Schema)
Agent streams data into a frontend-authored fixed component schema; backend ships JSON, agent fills in values
LangroidIn-Chat HITL (useInterrupt — low-level primitive)
Interactive component rendered inline in the chat via the lower-level `useInterrupt` primitive — direct control over the interrupt lifecycle
LangroidHeadless Interrupt (testing)
Resolve interrupts from a plain button grid — no chat, no useInterrupt render prop