# langchain-ai/agent-inbox

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

Repository: https://github.com/langchain-ai/agent-inbox
Canonical: https://ross.abutalabs.com/products/agent-inbox
Homepage: https://dev.agentinbox.ai
Language: TypeScript
License: MIT
License Family: permissive
Last push: 2026-09-03T00:31:52+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 35, longevity 47
- inputs: {"age_days": 667, "days_push": 0, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1086, forks 155 (observed 2026-09-03T02:15:14.274649+00:00)

## What it is
Agent Inbox is a web-based inbox UI for reviewing and responding to human-in-the-loop interrupts raised by LangGraph agents. It connects to any LangGraph deployment via a LangSmith API key and deployment URL, letting humans accept, edit, ignore, or respond to agent action requests.

## Use cases
- review and approve agent actions before execution
- add human-in-the-loop approval to LangGraph agents
- manage interrupts from multiple LangGraph deployments in one inbox
- edit or reject proposed agent tool calls
- self-host a UI for supervising AI agents
- test interrupt workflows in LangGraph Python or TypeScript agents

## When to choose
- you build agents with LangGraph and need a ready-made human oversight UI
- you want to self-host or use the hosted version to supervise agent interrupts
- you need accept/edit/respond/ignore actions on agent proposals

## When to avoid
- your agents are not built on LangGraph
- you need a fully custom human-review workflow beyond the interrupt schema
- you want a backend service rather than a browser-based UI

## Facets
- artifact type: application
- maturity: active
- function: chat-interface, ui-components, agent-framework, web-framework
- domain: large-language-models, developer-tools, web-development
- platform: self-hosted, cross-platform
- tags: human-in-the-loop, langgraph, langchain, agent-ux, inbox, interrupts, typescript, react, ai-agents, web-server, nodejs

## Member repositories
- langchain-ai/agent-inbox (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:14.274649+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:53:10.057486+00:00, confidence not recorded.
  - readme: https://github.com/langchain-ai/agent-inbox (fetched 2026-09-03T02:15:14.274649+00:00, sha 8326448279c9)
  - homepage: https://dev.agentinbox.ai (fetched 2026-08-29T12:54:14.501215+00:00, sha 430791100175)
- Data as of 2026-08-30T08:39:29.467469+00:00.