langchain-ai/agent-inbox

📥 An inbox UX for interacting with human-in-the-loop agents.

What it solves

Agent Inbox provides a user interface for managing "human-in-the-loop" interactions with AI agents. It solves the problem of how a human user can easily review, edit, and respond to interrupts triggered by a LangGraph agent when it needs human approval or input before proceeding.

How it works

The tool connects to a LangGraph deployment via an API key and deployment URL. When an agent triggers an interrupt function in its code, the Agent Inbox fetches these interrupts and displays them to the user. The user can then interact with the interrupt based on a predefined configuration:

  • Accept: Confirms the agent's proposed action or arguments.
  • Edit: Modifies the arguments of the action before the agent continues.
  • Response: Sends a custom text response back to the agent.
  • Ignore: Skips the interrupt.

Who it’s for

Developers building AI agents with LangGraph who need a professional interface for human oversight and interaction without building a custom frontend from scratch.

Highlights

  • LangGraph Integration: Specifically designed to work with LangGraph's interrupt system.
  • Flexible Response Types: Supports accepting, editing, responding to, or ignoring agent requests.
  • Customizable Interrupts: Developers can control which actions (ignore, respond, edit, accept) are available to the user for each specific interrupt.
  • Markdown Support: Allows detailed, markdown-formatted descriptions to provide context to the human reviewer.

Related

  • Project
    langchain-ai/langgraph

    A low-level orchestration framework for building, managing, and deploying long-running, stateful AI agents with durable execution and human-in-the-loop capabilities.

  • Project
    CopilotKit/open-research-ANA

    A research canvas application that combines real-time search and an agentic interface to simplify complex research tasks through human-in-the-loop interaction.

  • Project
    liangdabiao/langgraph_multi-agent-rag-customer-support

    A multi-agent RAG customer support system built with LangGraph and LangChain that features security guardrails and a human-in-the-loop approval process for sensitive operations.

  • Project
    langchain-ai/react-agent

    A LangGraph template for implementing a ReAct agent that iteratively reasons and executes tool-based actions to solve complex user queries.

  • Project
    langchain-ai/social-media-agent

    An LLM‑driven agent that scrapes a URL, drafts Twitter/LinkedIn posts, lets a human review them, and then publishes via Arcade or native OAuth. Built on LangGraph, it supports optional Slack, Supabase, GitHub, and YouTube integrations and is configurable through prompt files.