AG-UI
An interface streaming protocol for communication between agents and frontends. It describes what the interface should present, such as adding cards, updating table rows, appending text or triggering interactive regions, so that the frontend can render without understanding the agent's internal graph structure.
Timeline (last 90 days)
OpenBot supports connecting any agent that conforms to the AG-UI protocol and does not lock users into a single agent framework.
OpenBot is decoupled from upper-level agent frameworks via the AG-UI protocol, so developers can build agents with frameworks they are familiar with and then connect them to OpenBot to gain an independent computer and action auditing capabilities.
The AG-UI protocol was initiated and driven toward standardization by the CopilotKit team.
AG-UI and MCP are complementary: AG-UI focuses on the interaction layer between agents and user interfaces, while MCP focuses on connecting agents with tools/data.
LangGraph's graph state maps naturally to state update events in the event stream; Mastra's workflow steps align more closely with the granularity of tool call events.
OpenUI provides the @openui/langchain package, which is responsible for converting LangGraph streaming events into AG-UI events.
llama-index-protocols-ag-ui 0.5.0 no longer persists front-end tool results on the server side (#22727).
AG-UI is a standardized protocol proposed by LlamaIndex to regulate communication between AI Agent backends and front-end interfaces, covering interaction patterns such as tool calls, state synchronization, and streaming events.
Moka supports pointing to an existing LangGraph agent via the AG-UI protocol, serving as a debugging interface for that agent
The AG-UI (Agent-User Interaction) protocol is an emerging open standard for regulating real-time communication interfaces between front-end UI and back-end Agents, supporting scenarios such as streaming event push, tool call status synchronization, and human-in-the-loop interruption
All Facts (10)
OpenBot supports connecting any agent that conforms to the AG-UI protocol and does not lock users into a single agent framework.
50%UnverifiedOpenBot is decoupled from upper-level agent frameworks via the AG-UI protocol, so developers can build agents with frameworks they are familiar with and then connect them to OpenBot to gain an independent computer and action auditing capabilities.
50%UnverifiedThe AG-UI protocol was initiated and driven toward standardization by the CopilotKit team.
50%UnverifiedAG-UI and MCP are complementary: AG-UI focuses on the interaction layer between agents and user interfaces, while MCP focuses on connecting agents with tools/data.
50%UnverifiedLangGraph's graph state maps naturally to state update events in the event stream; Mastra's workflow steps align more closely with the granularity of tool call events.
50%UnverifiedOpenUI provides the @openui/langchain package, which is responsible for converting LangGraph streaming events into AG-UI events.
50%Unverifiedllama-index-protocols-ag-ui 0.5.0 no longer persists front-end tool results on the server side (#22727).
50%UnverifiedAG-UI is a standardized protocol proposed by LlamaIndex to regulate communication between AI Agent backends and front-end interfaces, covering interaction patterns such as tool calls, state synchronization, and streaming events.
50%UnverifiedMoka supports pointing to an existing LangGraph agent via the AG-UI protocol, serving as a debugging interface for that agent
50%UnverifiedThe AG-UI (Agent-User Interaction) protocol is an emerging open standard for regulating real-time communication interfaces between front-end UI and back-end Agents, supporting scenarios such as streaming event push, tool call status synchronization, and human-in-the-loop interruption
50%