# agentscope-ai/AgentTeams

An open-source Collaborative Multi-Agent OS for transparent, human-in-the-loop task coordination via Matrix rooms.

Repository: https://github.com/agentscope-ai/AgentTeams
Canonical: https://ross.abutalabs.com/products/agentteams
Homepage: https://www.aliyun.com/product/agentteams
Language: Go
License: Apache-2.0
License Family: permissive
Topics: agent-teams, openclaw
Last push: 2026-08-22T00:13:12+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 98, longevity 13
- inputs: {"age_days": 193, "days_push": 12, "days_rel": 12, "gap_med": 0, "n_releases_24m": 30}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5492, forks 674 (observed 2026-08-28T04:09:19.819058+00:00)

## What it is
AgentTeams is an open-source collaborative multi-agent runtime platform that orchestrates multiple Agent containers using a Manager-Workers architecture. It provides transparent, auditable, human-in-the-loop task coordination through Matrix rooms, with enterprise features like credential isolation, RBAC, and full-chain tracing.

## Use cases
- orchestrate multiple AI agents collaborating on tasks
- run a multi-agent team with human oversight and intervention
- govern enterprise agents with audit trails and access control
- mix different agent runtimes in one shared room
- isolate agent credentials so agents never see plaintext secrets
- monitor token usage and trace agent decisions
- deploy digital employees that work through company IM

## When to choose
- you need multiple heterogeneous agents to collaborate under centralized orchestration
- you require human-in-the-loop visibility and intervention in agent workflows
- you need enterprise-grade security, auditing, and RBAC for agent teams
- you want to manage existing agents without rewriting their logic
- you prefer IM-native (Matrix) collaboration between humans and agents

## When to avoid
- you need a framework that implements individual agent logic rather than orchestrating existing agents
- you want a lightweight single-agent setup without multi-agent coordination
- you don't need Matrix-based rooms or IM integration
- you need deep customization of agent reasoning rather than governance and orchestration

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, mcp, workflow-automation, monitoring, secrets-management, chatbot
- domain: large-language-models, developer-tools, self-hosted
- platform: go, self-hosted, cross-platform, cloud
- tags: multi-agent, human-in-the-loop, matrix, manager-workers, orchestration, enterprise, audit, rbac, credential-isolation, openclaw, ai-agents, automation, docker

## Member repositories
- agentscope-ai/AgentTeams (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:19.819058+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 2026-08-29T17:56:20.399738+00:00, confidence not recorded.
  - readme: https://github.com/agentscope-ai/AgentTeams (fetched 2026-08-28T04:09:19.819058+00:00, sha 9f4f34cc6211)
  - homepage: https://www.aliyun.com/product/agentteams (fetched 2026-08-29T08:51:44.491992+00:00, sha 2d416bafc90d)
  - site_page: https://www.aliyun.com/about (fetched 2026-08-29T08:51:44.501504+00:00, sha 2a1549f39cca)
  - site_page: https://www.aliyun.com/about/what-is-cloud-computing (fetched 2026-08-29T08:51:44.503530+00:00, sha e5f37aa83b38)
- Data as of 2026-08-30T08:39:29.467469+00:00.