agentscope-ai/AgentTeamsPublic

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

AI summary: An open-source Collaborative Multi-Agent OS orchestrating task coordination directly within Matrix rooms.

Stars
6K
+23 today
Forks
742
Watchers
28
Open issues
206
Open PRs
27
Contributors
~56
Commits
1K
Branches
230
Views (30d)
31

GoApache-2.0Created Feb 21, 2026Last push 3d agoLatest release v1.2.6+54 stars this week+262 this month

Quick answers

What is AgentTeams?
An open-source Collaborative Multi-Agent OS orchestrating task coordination directly within Matrix rooms.
What does AgentTeams do?
AgentTeams (hiclaw) is an open-source collaborative multi-agent runtime platform built on a scalable Manager-Workers architecture. It operates directly within an Instant Messaging environment using the Matrix protocol, allowing humans and multiple distinct AI agents (like OpenClaw and Hermes) to interact transparently in shared chat rooms. The platform does not implement internal agent logic itself, but instead orchestrates and manages multiple containerized agent runtimes. By leveraging a shared MinIO file system and centralizing traffic through the Higress AI Gateway, it significantly reduces redundant token consumption while maintaining strict enterprise security controls.
Who is AgentTeams for?
Platform engineers, AI developers, and enterprise IT teams building auditable multi-agent workflows with a need for strict Kubernetes deployment. Familiarity with multi-agent systems is highly recommended.
How do I get started with AgentTeams?
helm install agentteams agentscope/agentteams
How popular is AgentTeams on GitHub?
agentscope-ai/AgentTeams has 6,036 stars and 742 forks on GitHub, and gained 54 stars in the last 7 days.
What license does AgentTeams use?
agentscope-ai/AgentTeams is released under the Apache-2.0 license.

Star history

since Jul 29, 2026
02K4K6KJul 2026Sep 2026Oct 2026Nov 2026
6K stars as of Nov 18, 2026. Measured daily since Jul 29, 2026; GitHub no longer exposes earlier star timestamps.

Update history

3 recorded
  • Nov 10, 2026New release v1.2.6 (previously v1.2.5).
  • Oct 19, 2026New release v1.2.5 (previously v1.2.4).
  • Oct 4, 2026Previously tracked as higress-group/hiclaw; its 1 daily snapshot and 1 trending appearance were merged into this profile. Stars: 5,205 on 2026-07-29 under the old name, 5,696 on 2026-10-04 (+491).

Contribution activity

commits per day, last 52 weeks
NovDecJanFebMarAprMayJunJulAugSepOctNovMonWedFri2025-11-16: 0 commits2025-11-17: 0 commits2025-11-18: 0 commits2025-11-19: 0 commits2025-11-20: 0 commits2025-11-21: 0 commits2025-11-22: 0 commits2025-11-23: 0 commits2025-11-24: 0 commits2025-11-25: 0 commits2025-11-26: 0 commits2025-11-27: 0 commits2025-11-28: 0 commits2025-11-29: 0 commits2025-11-30: 0 commits2025-12-01: 0 commits2025-12-02: 0 commits2025-12-03: 0 commits2025-12-04: 0 commits2025-12-05: 0 commits2025-12-06: 0 commits2025-12-07: 0 commits2025-12-08: 0 commits2025-12-09: 0 commits2025-12-10: 0 commits2025-12-11: 0 commits2025-12-12: 0 commits2025-12-13: 0 commits2025-12-14: 0 commits2025-12-15: 0 commits2025-12-16: 0 commits2025-12-17: 0 commits2025-12-18: 0 commits2025-12-19: 0 commits2025-12-20: 0 commits2025-12-21: 0 commits2025-12-22: 0 commits2025-12-23: 0 commits2025-12-24: 0 commits2025-12-25: 0 commits2025-12-26: 0 commits2025-12-27: 0 commits2025-12-28: 0 commits2025-12-29: 0 commits2025-12-30: 0 commits2025-12-31: 0 commits2026-01-01: 0 commits2026-01-02: 0 commits2026-01-03: 0 commits2026-01-04: 0 commits2026-01-05: 0 commits2026-01-06: 0 commits2026-01-07: 0 commits2026-01-08: 0 commits2026-01-09: 0 commits2026-01-10: 0 commits2026-01-11: 0 commits2026-01-12: 0 commits2026-01-13: 0 commits2026-01-14: 0 commits2026-01-15: 0 commits2026-01-16: 0 commits2026-01-17: 0 commits2026-01-18: 0 commits2026-01-19: 0 commits2026-01-20: 0 commits2026-01-21: 0 commits2026-01-22: 0 commits2026-01-23: 0 commits2026-01-24: 0 commits2026-01-25: 0 commits2026-01-26: 0 commits2026-01-27: 0 commits2026-01-28: 0 commits2026-01-29: 0 commits2026-01-30: 0 commits2026-01-31: 0 commits2026-02-01: 0 commits2026-02-02: 0 commits2026-02-03: 0 commits2026-02-04: 0 commits2026-02-05: 0 commits2026-02-06: 0 commits2026-02-07: 0 commits2026-02-08: 0 commits2026-02-09: 0 commits2026-02-10: 0 commits2026-02-11: 0 commits2026-02-12: 0 commits2026-02-13: 3 commits2026-02-14: 0 commits2026-02-15: 0 commits2026-02-16: 0 commits2026-02-17: 2 commits2026-02-18: 2 commits2026-02-19: 2 commits2026-02-20: 4 commits2026-02-21: 13 commits2026-02-22: 0 commits2026-02-23: 2 commits2026-02-24: 4 commits2026-02-25: 1 commit2026-02-26: 2 commits2026-02-27: 12 commits2026-02-28: 6 commits2026-03-01: 6 commits2026-03-02: 10 commits2026-03-03: 14 commits2026-03-04: 25 commits2026-03-05: 36 commits2026-03-06: 14 commits2026-03-07: 6 commits2026-03-08: 6 commits2026-03-09: 19 commits2026-03-10: 12 commits2026-03-11: 18 commits2026-03-12: 6 commits2026-03-13: 21 commits2026-03-14: 37 commits2026-03-15: 33 commits2026-03-16: 13 commits2026-03-17: 11 commits2026-03-18: 16 commits2026-03-19: 5 commits2026-03-20: 13 commits2026-03-21: 14 commits2026-03-22: 4 commits2026-03-23: 3 commits2026-03-24: 9 commits2026-03-25: 5 commits2026-03-26: 0 commits2026-03-27: 8 commits2026-03-28: 9 commits2026-03-29: 10 commits2026-03-30: 3 commits2026-03-31: 5 commits2026-04-01: 7 commits2026-04-02: 9 commits2026-04-03: 4 commits2026-04-04: 0 commits2026-04-05: 0 commits2026-04-06: 0 commits2026-04-07: 3 commits2026-04-08: 11 commits2026-04-09: 0 commits2026-04-10: 0 commits2026-04-11: 0 commits2026-04-12: 0 commits2026-04-13: 1 commit2026-04-14: 4 commits2026-04-15: 1 commit2026-04-16: 11 commits2026-04-17: 1 commit2026-04-18: 8 commits2026-04-19: 2 commits2026-04-20: 4 commits2026-04-21: 7 commits2026-04-22: 12 commits2026-04-23: 14 commits2026-04-24: 15 commits2026-04-25: 1 commit2026-04-26: 7 commits2026-04-27: 2 commits2026-04-28: 3 commits2026-04-29: 0 commits2026-04-30: 0 commits2026-05-01: 0 commits2026-05-02: 0 commits2026-05-03: 0 commits2026-05-04: 0 commits2026-05-05: 0 commits2026-05-06: 11 commits2026-05-07: 6 commits2026-05-08: 0 commits2026-05-09: 0 commits2026-05-10: 0 commits2026-05-11: 8 commits2026-05-12: 1 commit2026-05-13: 0 commits2026-05-14: 0 commits2026-05-15: 0 commits2026-05-16: 0 commits2026-05-17: 0 commits2026-05-18: 0 commits2026-05-19: 4 commits2026-05-20: 9 commits2026-05-21: 13 commits2026-05-22: 3 commits2026-05-23: 6 commits2026-05-24: 0 commits2026-05-25: 1 commit2026-05-26: 1 commit2026-05-27: 3 commits2026-05-28: 3 commits2026-05-29: 3 commits2026-05-30: 0 commits2026-05-31: 0 commits2026-06-01: 8 commits2026-06-02: 2 commits2026-06-03: 1 commit2026-06-04: 1 commit2026-06-05: 0 commits2026-06-06: 0 commits2026-06-07: 3 commits2026-06-08: 1 commit2026-06-09: 1 commit2026-06-10: 2 commits2026-06-11: 0 commits2026-06-12: 0 commits2026-06-13: 0 commits2026-06-14: 0 commits2026-06-15: 1 commit2026-06-16: 5 commits2026-06-17: 0 commits2026-06-18: 0 commits2026-06-19: 0 commits2026-06-20: 0 commits2026-06-21: 1 commit2026-06-22: 1 commit2026-06-23: 1 commit2026-06-24: 0 commits2026-06-25: 0 commits2026-06-26: 0 commits2026-06-27: 1 commit2026-06-28: 0 commits2026-06-29: 0 commits2026-06-30: 0 commits2026-07-01: 0 commits2026-07-02: 2 commits2026-07-03: 1 commit2026-07-04: 5 commits2026-07-05: 7 commits2026-07-06: 1 commit2026-07-07: 0 commits2026-07-08: 3 commits2026-07-09: 2 commits2026-07-10: 0 commits2026-07-11: 0 commits2026-07-12: 5 commits2026-07-13: 0 commits2026-07-14: 1 commit2026-07-15: 1 commit2026-07-16: 0 commits2026-07-17: 7 commits2026-07-18: 8 commits2026-07-19: 0 commits2026-07-20: 1 commit2026-07-21: 2 commits2026-07-22: 0 commits2026-07-23: 0 commits2026-07-24: 1 commit2026-07-25: 1 commit2026-07-26: 2 commits2026-07-27: 0 commits2026-07-28: 1 commit2026-07-29: 3 commits2026-07-30: 5 commits2026-07-31: 2 commits2026-08-01: 6 commits2026-08-02: 0 commits2026-08-03: 1 commit2026-08-04: 0 commits2026-08-05: 0 commits2026-08-06: 4 commits2026-08-07: 1 commit2026-08-08: 5 commits2026-08-09: 1 commit2026-08-10: 0 commits2026-08-11: 1 commit2026-08-12: 0 commits2026-08-13: 0 commits2026-08-14: 3 commits2026-08-15: 6 commits2026-08-16: 0 commits2026-08-17: 0 commits2026-08-18: 1 commit2026-08-19: 4 commits2026-08-20: 1 commit2026-08-21: 2 commits2026-08-22: 1 commit2026-08-23: 0 commits2026-08-24: 0 commits2026-08-25: 0 commits2026-08-26: 0 commits2026-08-27: 0 commits2026-08-28: 0 commits2026-08-29: 0 commits2026-08-30: 0 commits2026-08-31: 0 commits2026-09-01: 0 commits2026-09-02: 1 commit2026-09-03: 0 commits2026-09-04: 2 commits2026-09-05: 2 commits2026-09-06: 0 commits2026-09-07: 0 commits2026-09-08: 0 commits2026-09-09: 0 commits2026-09-10: 2 commits2026-09-11: 5 commits2026-09-12: 1 commit2026-09-13: 0 commits2026-09-14: 2 commits2026-09-15: 6 commits2026-09-16: 19 commits2026-09-17: 2 commits2026-09-18: 12 commits2026-09-19: 4 commits2026-09-20: 7 commits2026-09-21: 6 commits2026-09-22: 0 commits2026-09-23: 3 commits2026-09-24: 0 commits2026-09-25: 0 commits2026-09-26: 0 commits2026-09-27: 0 commits2026-09-28: 0 commits2026-09-29: 0 commits2026-09-30: 0 commits2026-10-01: 0 commits2026-10-02: 0 commits2026-10-03: 8 commits2026-10-04: 4 commits2026-10-05: 0 commits2026-10-06: 0 commits2026-10-07: 0 commits2026-10-08: 8 commits2026-10-09: 7 commits2026-10-10: 3 commits2026-10-11: 4 commits2026-10-12: 1 commit2026-10-13: 4 commits2026-10-14: 2 commits2026-10-15: 4 commits2026-10-16: 2 commits2026-10-17: 1 commit2026-10-18: 0 commits2026-10-19: 0 commits2026-10-20: 0 commits2026-10-21: 0 commits2026-10-22: 0 commits2026-10-23: 7 commits2026-10-24: 6 commits2026-10-25: 0 commits2026-10-26: 0 commits2026-10-27: 5 commits2026-10-28: 2 commits2026-10-29: 10 commits2026-10-30: 4 commits2026-10-31: 5 commits2026-11-01: 7 commits2026-11-02: 9 commits2026-11-03: 3 commits2026-11-04: 5 commits2026-11-05: 5 commits2026-11-06: 26 commits2026-11-07: 11 commits2026-11-08: 15 commits2026-11-09: 16 commits2026-11-10: 12 commits2026-11-11: 1 commit2026-11-12: 0 commits2026-11-13: 0 commits2026-11-14: 0 commits
1,041 commits in the last yearLessMore

Signals and awards

derived from tracked data
  • Very active

    1,041 commits in 52 weeks

  • Permissive license

    Apache-2.0

  • Continuous integration

    Automated checks passing

What AgentTeams does

AgentTeams (hiclaw) is an open-source collaborative multi-agent runtime platform built on a scalable Manager-Workers architecture. It operates directly within an Instant Messaging environment using the Matrix protocol, allowing humans and multiple distinct AI agents (like OpenClaw and Hermes) to interact transparently in shared chat rooms. The platform does not implement internal agent logic itself, but instead orchestrates and manages multiple containerized agent runtimes. By leveraging a shared MinIO file system and centralizing traffic through the Higress AI Gateway, it significantly reduces redundant token consumption while maintaining strict enterprise security controls.

Platform engineers, AI developers, and enterprise IT teams building auditable multi-agent workflows with a need for strict Kubernetes deployment. Familiarity with multi-agent systems is highly recommended.

  • Manager-Workers Architecture: Orchestrates multiple specialized worker agents through a central manager, reducing the need for continuous human oversight.
  • Multi-Runtime Collaboration: Supports executing deterministic agents alongside autonomous LLM agents within the same collaborative session.
  • Matrix Protocol Integration: Operates natively within standard IM clients like Element, avoiding heavy enterprise UI integration overhead.
  • Shared File System: Utilizes MinIO for inter-agent information exchange to drastically minimize redundant context window token consumption.
  • AI Gateway Routing: Employs Higress to centralize external traffic management and mitigate credential-related security vulnerabilities.

Where teams use it

Complex task orchestration

Coordinates specialized agents to tackle multi-step enterprise workflows requiring different underlying models and deterministic tools.

Human-in-the-loop operations

Provides a transparent chat interface where human operators can monitor, pause, or redirect agent teams mid-execution.

Enterprise AI governance

Centralizes API credentials and routing policies through a secure gateway to prevent token leakage and enforce usage limits.

Cross-platform agent integration

Bridges disparate agent frameworks into a unified execution environment without requiring developers to rewrite core logic.

Getting started: helm install agentteams agentscope/agentteams

README

main branch

AgentTeams

English | 中文 | 日本語

DeepWiki Discord

AgentTeams is an open-source collaborative multi-agent runtime platform. It enables multiple Agents to collaborate in a controlled and auditable room, with full human visibility and intervention capabilities throughout the process.

Built on a Manager-Workers architecture, AgentTeams features a Manager that centrally orchestrates multiple Workers, focusing on collaboration scenarios between humans and Agents, as well as among Agents within enterprise environments.

AgentTeams does not compete with other Agent runtimes. Instead of implementing Agent logic itself, it orchestrates and manages multiple Agent containers (including the Manager and numerous Workers).

Key Features

  • 🧬 Manager-Workers Architecture: Eliminates the need for human oversight of individual Worker Claws by enabling Agents to manage other Agents.

  • 🤝 Multi-Runtime Collaboration: OpenClaw, QwenPaw, Hermes, and experimental DeepSeek Harness Workers coexist in the same IM room. Use the runtime that best fits each role while keeping collaboration visible in Matrix.

  • 📦 MinIO Shared File System: Introduces a shared file system for inter-Agent information exchange, significantly reducing token consumption in multi-Agent collaboration scenarios.

  • 🔐 Higress AI Gateway: Centralizes traffic management and mitigates credential-related risks, alleviating user concerns about security vulnerabilities in the native Lobster framework.

  • ☎️ Element IM Client + Tuwunel IM Server (both Matrix protocol-based): Eliminating DingTalk/Lark integration overhead and enterprise approval workflows. Enables rapid user onboarding to experience the "delight" of model services within an IM environment, while maintaining compatibility with native OpenClaw IM integration.

  • 🧬 Integrate AgentLoop :Provides capabilities such as full-stack Agent observability and auditing, Agent evaluation and experimentation, and Agent asset management and continuous optimization.

News

  • 2026-11-08: Release Notes — AgentTeams v1.2.6: pairs with Dashboard v1.4.0 for a refreshed workspace, voice input, and optional QwenPaw/Hermes session tracing; adds task pause/resume controls and a team attention inbox; strengthens multi-user permissions and model grants; expands runtime configuration, human file areas, and S3-compatible storage; updates the experimental DeepSeek Harness Worker image to v0.1.1; and improves Matrix collaboration and upgrade reliability.
  • 2026-10-16: Release Notes — AgentTeams v1.2.5: strengthens coding-tool delegation, Worker configuration management, and in-place QwenPaw runtime upgrades; ships with Dashboard v1.3.1 by default; and fixes multiple issues in MCP registration, Matrix message collaboration, and Worker recovery.
  • 2026-09-20: Release Notes — AgentTeams v1.2.4: adds TeamHarness task state history, progress tracking, and completion notifications; upgrades to QwenPaw 2.2.1 and Dashboard v1.2.4.9; expands team permissions, auditing, and skill management; introduces the experimental DeepSeek Harness Worker; and improves Worker state persistence and installation reliability.
  • 2026-08-22: Release Notes — AgentTeams v1.2.3: makes long-running Project workflows visible and controllable through new Controller APIs, agt commands, and Dashboard v1.2.4; adds audited human interventions, Project history, and QwenPaw Worker checkpoints; promotes QwenPaw as the recommended local Manager and Worker runtime; and improves Skill recovery, Windows AppService startup, and runtime reliability.
  • 2026-08-08: Release Notes — AgentTeams v1.2.2: adds Manager-to-Worker custom Skill delivery with validation, storage upload, Worker.spec.skills assignment, and QwenPaw hot refresh and enablement without a restart; hardens Manager and Worker Skill loading; and ensures Team Leaders and Workers explicitly join Team Rooms after invitation.
  • 2026-08-06: Release Notes — AgentTeams v1.2.1: unifies the Manager and Worker runtime stack on QwenPaw 2.0.1, publishes the QwenPaw Worker multi-architecture image, strengthens atomic Matrix and TeamHarness task assignment and CoPaw-to-QwenPaw state migration, adds custom model vision and reasoning capability overrides, and improves Worker lifecycle and runtime management reliability.
  • 2026-07-30: Release Notes — AgentTeams v1.2.0 (stable): establishes AgentTeams naming and the final Team/Worker resource contracts end to end; adds the optional AgentTeams Dashboard; and improves Worker storage sync, Team routing and lifecycle convergence, installer support for deploying v1.1.2 with its legacy environment and storage contract (earlier releases still require the matching legacy installer), Dashboard reliability, and tooling and diagnostic safety.
  • 2026-07-17: Release Notes — AgentTeams v1.2.0-beta.1 (prerelease): completes the public rename from the retired predecessor across images, Kubernetes APIs, Helm, Matrix, storage, and runtime contracts; adds the plugin platform, TeamHarness and WorkerFlow integrations, Matrix AppService and Human SSO, model-provider routing and LLM preflight, plus richer controller observability. Beta installation requires explicit opt-in, while the stable default remains v1.1.2.
  • 2026-05-27: Release Notes — AgentTeams v1.1.2: QwenPaw-first installer with keep-all upgrade flow, Team human coordinators and refreshed Team Leader coordination tools, Nacos remote skills with sts-agentteams / ai-registry STS scope, Worker CR-name decoupled from runtime name, controller reconcile metrics and graceful shutdown.
  • 2026-05-07: Release Notes | Changelog — AgentTeams v1.1.1: declarative MCP on Worker/Manager/Team CRDs (breaking) and on Team Leader, custom spec.env for CRs, Token Plan + Qwen Cloud international + qwen3.6-plus, namespace-scoped controller RBAC, optional SOUL.md in Worker packages.
  • 2026-04-24: English | 中文 — AgentTeams v1.1.0: Kubernetes-native control plane, Hermes autonomous coding agent runtime, 1.7 GB image shrink, agt CLI replaces shell scripts.
  • 2026-04-14: English | 中文 — Deep dive: AgentTeams as a Kubernetes-native multi-agent collaboration orchestration system.
  • 2026-04-03: English | 中文 — AgentTeams 1.0.9: Kubernetes-style declarative resource management (YAML for Worker, Team, Human); Worker Template Marketplace; Manager QwenPaw runtime; Nacos Skills Registry and more.
  • 2026-03-14: English | 中文 — AgentTeams 1.0.6: enterprise-grade MCP Server management, zero credential exposure.
  • 2026-03-10: English | 中文 — AgentTeams 1.0.4: QwenPaw (formerly CoPaw) Worker support, 80% less memory.
  • 2026-03-04: English | 中文 — AgentTeams open sourced under its former name.

Why AgentTeams

  • Enterprise-Grade Security: Worker Agents operate with consumer tokens only. Real credentials (API keys, GitHub PATs) stay in the gateway — Workers can't see them, and neither can attackers.

  • Fully Private: Matrix is a decentralized, open protocol. Host it yourself, federate with others if you want. No vendor lock-in, no data harvesting.

  • Human-in-the-Loop by Default: Every Matrix room includes you, the Manager, and the relevant Workers. Watch everything. Jump in anytime. No black boxes.

  • Zero Configuration IM: Built-in Matrix server means no bot applications, no API approvals, no waiting. Just open Element Web and start chatting.

  • One Command Setup: curl | bash and you're done — AI gateway, Matrix server, file storage, web client, and Manager Agent.

  • Skills Ecosystem: Workers pull from skills.sh (80,000+ community skills) on demand. Safe because Workers can't access real credentials.

Quick Start

Prerequisites: OrbStack, Docker Desktop, Docker Engine, or compatible Podman. If no engine is ready, the installer guides you to install/start one and waits before continuing.

Resources: 2 CPU cores + 4 GB RAM minimum. For multiple Workers, 4 cores + 8 GB recommended.

Install

macOS / Linux:

bash <(curl -sSL https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.sh)

Windows (PowerShell 7+ recommended):

Set-ExecutionPolicy Bypass -Scope Process -Force; $wc=New-Object Net.WebClient; $wc.Encoding=[Text.Encoding]::UTF8; iex $wc.DownloadString('https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.ps1')

The default command starts the local services and Dashboard without configuration questions. API keys, embedding and Worker runtimes can be configured later. For the interactive setup, add --manual (PowerShell: -Manual).

Access

Dashboard opens automatically at http://127.0.0.1:13000 with admin prefilled and the generated password copied to your clipboard. Paste the password and sign in; if the browser or clipboard is unavailable, use the address and credentials printed in the terminal.

If no model is configured, use the Manager conversation's configuration link to open model management, enter your API key and choose a model. An exported AGENTTEAMS_LLM_API_KEY is preserved. Ask the Manager to create your first Worker and choose its runtime then; Worker images download on demand. See the quickstart for details.

Mobile: Use any Matrix client (Element, FluffyChat) and connect to your server address.

That's it. No bot applications. No external services. Your AI team runs entirely on your machine.

Upgrade

Run the same command again and confirm once to keep the original accounts, model settings, data and workspaces. Images are prepared before services are replaced, and the original configuration is backed up. See upgrade instructions for recovery and database backup/rollback options.

# Upgrade to latest (preserves all data)
bash <(curl -sSL https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.sh)

# Upgrade to specific version
AGENTTEAMS_VERSION=v1.2.5 bash <(curl -sSL https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.sh)

Uninstall

macOS / Linux:

bash <(curl -fsSL https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.sh) uninstall

Windows (PowerShell):

Set-ExecutionPolicy Bypass -Scope Process -Force; $wc=New-Object Net.WebClient; $wc.Encoding=[Text.Encoding]::UTF8; $s=$wc.DownloadString('https://raw.githubusercontent.com/agentscope-ai/AgentTeams/main/install/agentteams-install.ps1'); & ([scriptblock]::Create($s)) uninstall

This removes all AgentTeams containers (Manager, Workers, docker-proxy), Docker volume, network, env file, workspace directory, and install log.

Install on Kubernetes (Helm)

For shared / production deployments you can install AgentTeams on any Kubernetes cluster via the official Helm chart. The default profile bundles the Higress AI gateway, Tuwunel (Matrix), MinIO and the AgentTeams controller — no external dependencies required.

The commands below provide a quick installation path. See the Kubernetes Deployment Guide for cluster planning, values files, model services, runtimes, persistence, Ingress, and operations.

Prerequisites

  • Kubernetes 1.24+ (kind / minikube / k3s / managed K8s — all work)
  • Helm 3.7+
  • A default StorageClass (for the Tuwunel + MinIO PVCs)

Install (OpenAI / OpenAI-compatible)

helm repo add higress.io https://higress.io/helm-charts
helm repo update

helm install agentteams higress.io/agentteams \
  -n agentteams-system --create-namespace \
  --render-subchart-notes \
  --set credentials.llmApiKey=<your-api-key> \
  --set credentials.adminPassword=<your-admin-password> \
  --set gateway.publicURL=http://localhost:18080

For non-OpenAI providers that expose an OpenAI-compatible API, also set llmBaseUrl:

helm install agentteams higress.io/agentteams \
  -n agentteams-system --create-namespace \
  --render-subchart-notes \
  --set credentials.llmApiKey=<your-api-key> \
  --set credentials.llmBaseUrl=https://your-provider.example.com/v1 \
  --set credentials.defaultModel=your-model-name \
  --set credentials.adminPassword=<your-admin-password> \
  --set gateway.publicURL=http://localhost:18080
Using Qwen (通义千问) instead
helm install agentteams higress.io/agentteams \
  -n agentteams-system --create-namespace \
  --render-subchart-notes \
  --set credentials.llmApiKey=<your-qwen-api-key> \
  --set credentials.llmProvider=qwen \
  --set credentials.defaultModel=qwen3.5-plus \
  --set credentials.adminPassword=<your-admin-password> \
  --set gateway.publicURL=http://localhost:18080
Value Required Description
credentials.llmApiKey yes API key for your LLM provider
gateway.publicURL yes Public URL where users will reach Element Web (e.g. http://localhost:18080 for port-forward, or https://agentteams.example.com for an Ingress)
credentials.adminPassword recommended Matrix admin password; auto-generated if left empty (you'll have to read it back from the Secret)
credentials.llmProvider no LLM provider name, defaults to openai-compat
credentials.defaultModel no Default model, defaults to gpt-5.4
credentials.llmBaseUrl no OpenAI-compatible base URL (e.g. https://api.deepseek.com/v1). Leave empty for official OpenAI API
preflight.llm.enabled no Run an install/upgrade hook that validates the LLM API key, base URL, and model before the controller starts. Defaults to true
preflight.llm.strict no Fail the Helm install/upgrade when the LLM preflight fails. Defaults to true; set to false to emit a warning and continue
preflight.llm.timeoutSeconds no Per-request timeout for the LLM preflight HTTP probe. Defaults to 30
preflight.llm.retries no Retry count for transient LLM preflight failures such as rate limits and provider 5xx responses. Defaults to 2
preflight.llm.activeDeadlineSeconds no Kubernetes Job active deadline for the preflight hook. Defaults to 120
preflight.llm.resources no Optional Kubernetes resource requests/limits for the preflight hook container
manager.runtime no Manager runtime: OpenClaw uses openclaw (default); CoPaw uses qwenpaw in the current chart, with copaw retained as a compatibility alias. Managers do not support Hermes or DeepSeek Harness
worker.defaultRuntime no The chart provides default image values for openclaw (default), copaw, hermes, experimental deepseek-harness, and openhuman; the current CRD rejects an explicit spec.runtime: openhuman, and QwenPaw Workers require an explicit spec.image

Helm installs run an LLM preflight hook by default. The hook sends a minimal OpenAI-compatible /chat/completions request using credentials.llmApiKey, credentials.llmBaseUrl, and credentials.defaultModel; invalid keys, unreachable base URLs, unsupported models, quota errors, and provider outages fail the install before the controller starts. To bypass this check for restricted or offline clusters:

helm install agentteams higress.io/agentteams \
  -n agentteams-system --create-namespace \
  --set credentials.llmApiKey=<your-api-key> \
  --set credentials.adminPassword=<your-admin-password> \
  --set gateway.publicURL=http://localhost:18080 \
  --set preflight.llm.enabled=false
Using alternative runtimes (CoPaw Manager + Hermes Workers)
helm install agentteams higress.io/agentteams \
  -n agentteams-system --create-namespace --devel \
  --set manager.runtime=qwenpaw \
  --set manager.image.repository=higress-registry.cn-hangzhou.cr.aliyuncs.com/agentteams/agentteams-manager-qwenpaw \
  --set worker.defaultRuntime=hermes \
  --set credentials.llmApiKey=<your-api-key> \
  --set credentials.llmBaseUrl=https://your-provider.example.com/v1 \
  --set credentials.defaultModel=your-model-name \
  --set credentials.adminPassword=<your-admin-password> \
  --set gateway.publicURL=http://localhost:18080

When selecting the CoPaw Manager, set both manager.runtime=qwenpaw and the agentteams-manager-qwenpaw image. Configure Worker default images through worker.defaultImage.<runtime>.

Multi-Region Image Registry

Images point to the China region by default. When deploying outside China, use a nearby registry for faster pulls:

Region Registry
China (default) higress-registry.cn-hangzhou.cr.aliyuncs.com/higress
North America higress-registry.us-west-1.cr.aliyuncs.com/higress
Southeast Asia higress-registry.ap-southeast-7.cr.aliyuncs.com/higress

global.imageRegistry affects only subcharts that consume that global value. The Controller, Manager, Workers, Tuwunel, MinIO, and Element Web use their own complete image.repository values. Override every relevant image in a values file when switching regions or using a private registry. See the Kubernetes Deployment Guide and helm/agentteams/values.yaml for the complete configuration.

Access

For a temporary local admin session, forward the Higress Gateway:

kubectl port-forward -n agentteams-system svc/higress-gateway 18080:80

Then open http://localhost:18080 and log in to Element Web. The port-forward ends when the command exits and is not suitable for shared access.

For company or Internet access, expose only svc/higress-gateway through an HTTPS Ingress or LoadBalancer. gateway.publicURL is written into the Element Web configuration as its Matrix homeserver URL, so it must exactly match the public origin that users open (for example, https://agentteams.example.com).

  1. Point the public DNS name at your Ingress controller or load balancer.
  2. Provision a TLS certificate Secret in agentteams-system.
  3. Set the same origin on the Helm release:
helm upgrade agentteams higress.io/agentteams \
  -n agentteams-system --reuse-values \
  --set gateway.publicURL=https://agentteams.example.com
  1. Route that host to the Higress Gateway. This generic example assumes an NGINX IngressClass and an existing agentteams-tls Secret; replace both to match your cluster:
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: agentteams
  namespace: agentteams-system
spec:
  ingressClassName: nginx
  tls:
    - hosts:
        - agentteams.example.com
      secretName: agentteams-tls
  rules:
    - host: agentteams.example.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: higress-gateway
                port:
                  number: 80

Verify both the web entry point and Matrix routing after DNS and TLS are ready:

curl -fsSI https://agentteams.example.com/
curl -fsS https://agentteams.example.com/_matrix/client/versions

Keep the controller API, Tuwunel, MinIO, and the Higress Console private unless you apply separate authentication and network policy. Use HTTPS for shared access because Matrix login credentials and access tokens pass through this endpoint. A LoadBalancer Service can replace the Ingress, but the DNS, TLS, and gateway.publicURL requirements remain the same.

Upgrade

helm repo update
helm upgrade agentteams higress.io/agentteams -n agentteams-system --reuse-values

Uninstall

helm uninstall agentteams -n agentteams-system
kubectl delete namespace agentteams-system

For the Kubernetes-native architecture (CRDs, controller, declarative Worker / Team / Human resources), see docs/design/k8s-native-orchestration.md.

How It Works

Manager as Your AI Chief of Staff

You: Create a Worker named alice for frontend development

Manager: Done. Worker alice is ready.
         Room: Worker: Alice
         Tell alice what to build.

You: @alice implement a login page with React

Alice: On it... [a few minutes later]
       Done. PR submitted: https://github.com/xxx/pull/1

    

① Manager creates a Worker and assigns tasks                    ② You can also direct Workers directly in the room

Security Model

Worker (consumer token only)
    → Higress AI Gateway (holds real API keys, GitHub PAT)
        → LLM API / GitHub API / MCP Servers

Workers see only their consumer token. The gateway handles all real credentials. The Manager knows what Workers are doing but never touches the actual keys.

Human in the Loop

Every Matrix Room includes you, the Manager, and relevant Workers:

You: @bob wait, change the password rule to minimum 8 chars
Bob: Got it, updated.
Alice: Frontend validation updated too.

No hidden agent-to-agent calls. Everything is visible and intervenable.

Multi-Runtime Collaboration

AgentTeams provides multiple Worker runtime families that can coexist in the same IM room, collaborating on tasks together:

  • OpenClaw (Node.js) — General-purpose agent with rich skills ecosystem, ideal for task orchestration and tool calling
  • QwenPaw (Python) — Lightweight runtime, suited for browser automation and quick tasks
  • Hermes (hermes-agent) — Autonomous coding agent with terminal sandbox, self-improving skills, and persistent memory
  • DeepSeek Harness (experimental) — Headless coding agent with room-scoped session continuation, Matrix file exchange, and restart recovery; currently pinned to a tested DSH release candidate

The copaw value remains available for compatibility with existing CoPaw Workers; the current CRD accepts both copaw and qwenpaw. Each runtime excels at different tasks. A common pattern is to use deterministic agents (OpenClaw/QwenPaw) as Leaders to decompose and assign work, and Hermes Workers for autonomous code execution. All runtimes communicate through Matrix rooms—fully visible and fully intervenable.

# Switch any worker's runtime in place
agt update worker --runtime hermes

Architecture

┌───────────────────────────────────────────────┐
│            agentteams-controller                  │
│  Higress │ Tuwunel │ MinIO │ Element Web      │
└──────────────────┬────────────────────────────┘
                   │ Matrix + HTTP Files
┌──────────────────┴──────────┐
│     agentteams-manager-agent     │
│     Manager (OpenClaw/       │
│       QwenPaw)               │
└──────────────────┬──────────┘
                   │
┌──────────────────┼────────────────────────────┐
│                  │                            │
▼                  ▼                            ▼
Worker Alice    Worker Bob              Worker Charlie
(OpenClaw)      (QwenPaw)               (Hermes)
Component Role
agentteams-controller Kubernetes-native control plane, reconciles Worker/Team/Manager CRs
Higress AI Gateway LLM proxy, MCP Server hosting, credential management
Tuwunel (Matrix) Self-hosted IM server for all Agent + Human communication
Element Web Browser client, zero setup
MinIO Centralized file storage, Workers are stateless

AgentTeams vs OpenClaw Native

OpenClaw Native AgentTeams
Deployment Single process Distributed containers
Agent creation Manual config + restart Conversational
Credentials Each agent holds real keys Workers only hold consumer tokens
Human visibility Optional Built-in (Matrix Rooms)
Mobile access Depends on channel setup Any Matrix client, zero config
Monitoring None Manager heartbeat, visible in Room

Documentation

Browse the documentation directory, or start with the overview and quickstart:

docs/overview.md Product overview, core concepts, and documentation map
docs/quickstart.md Step-by-step guide
docs/usage/use-cases.md Software delivery, research, content, incident analysis, and long-running project collaboration examples
docs/usage/deployment/local.md Local instance creation, installation options, upgrades, and uninstalling
docs/design/architecture.md System architecture deep dive (also available in 中文)
docs/design/k8s-native-orchestration.md Kubernetes-native multi-Agent orchestration: CRDs, controller, declarative resources (also available in 中文)
docs/usage/manager-guide.md Manager configuration
docs/usage/worker-guide.md Worker deployment
docs/usage/single-and-multi-user.md One deployment for one person or many: roles, permission levels, and what changes when you add people (also available in 中文)
docs/usage/project-workflow-api.md Project / workflow inspection API (also available in 中文)
docs/usage/module-update.md Module update & distribution: Millforge channels, release index, and the three install planes (also available in 中文)
docs/usage/higress-gateway-api.md Higress gateway external API reference (LLM / MCP / exposed ports / Console) (also available in 中文)
docs/usage/development.md Contributing and local dev (also available in 中文)

The remaining design documents (skill catalog, worker channels/chats APIs, room power levels, capability foundation, human update / audit / AI-routes APIs, L2/L3 scoped access, member runtime config contract, team skills, agent pod template, TeamHarness v0.1 boundary) are indexed in the 中文 documentation table — all are also available under docs/zh-cn/design/, except a few (docs/design/internal/, part of docs/design/teamharness/) that are natively written in Chinese upstream.

Troubleshooting

docker exec -it agentteams-manager cat /var/log/agentteams/manager-agent.log

See docs/usage/troubleshooting/faq.md for common issues (also available in 中文).

Reporting Bugs

Export your Matrix message logs and let an AI tool analyze them against the codebase before filing an issue — this helps us fix bugs much faster.

# Export debug logs (Matrix messages + agent sessions, PII auto-redacted)
python scripts/export-debug-log.py --range 1h

Then open the AgentTeams repo in Cursor, Claude Code, or similar AI tool and ask:

"Read the JSONL files in debug-log/. Analyze the Matrix message logs and agent session logs together. Cross-reference with the AgentTeams codebase to identify the root cause of [describe your bug]."

Include the AI's analysis in your bug report.

You can also let the AI tool submit the issue or PR directly. Install GitHub CLI, run gh auth login to authenticate in your browser, then add the OpenClaw GitHub skill to your AI coding tool (Cursor, Claude Code, etc.). After that, just ask it to file the issue or open a PR based on its analysis.

Build & Test

make build          # Build all images
make test           # Build + run all integration tests
make test-quick     # Smoke test only

Other Commands

make replay TASK="Create a Worker named alice for frontend development"
make uninstall
make help

Community

License

Apache License 2.0

View on GitHub

Recent activity

commits and pull requests

Releases and announcements

34 total
  1. v1.2.6v1.2.6Nov 8, 202628 downloads

    # AgentTeams 1.2.6 Release Notes AgentTeams 1.2.6 重点增强多 Agent 协作的过程可见性、任务控制和多用户权限管理,默认搭配 Dashboard v1.4.0。新版工作台加入语音输入与执行追踪,并完善运行时配置、文件存储和升级流程;实验性 DeepSeek Harness Worker 镜像更新至 v0.1.1。 ## 新增与增强 ### Dashboard v1.4.0:更集中的工作台与语音输入 - 将常用功能归为「助手、工作台、资产中心、成员与团队」四个主菜单,改善多团队会话切换与任务查看体验。 - 聊天新增语音转文字,默认使用浏览器本地免费识别;识别结果先进入输入框,可编辑后手动发送。管理员也可配置百炼语音识别。 - 支持中英文语音输入,改进中文转写结果处理;本地模型首次使用需要加载,之后会缓存。 - 优化消息时间与正文的排版、已读状态持久化,并新增中性主题及可选的日夜切换。 ### 原生会话追踪与执行诊断 - 新增可选的 Worker 原生会话追踪,支持关联 QwenPaw、Hermes 的 turn、模型请求和工具调用。 - Dashboard 增加按用户权限过滤的会话列表与执行瀑布图,便于查看耗时分布、调用细节及失败步骤。 - 补充 Tracing 后端配置与 Manager 配置指引,并修正追踪记录的团队身份归属。 - Matrix 增加静默丢弃消息的工作区记录,改善「消息已发出但 Agent 没有响应」时的排查线索。 ### 任务控制、人工介入与协作记录 - TeamHarness 支持暂停、恢复任务派发与 Loop 推进,并记录 Loop 预算和控制操作审计。 - 新增团队级人工事项收件箱,提供项目 attention 查询、结构化人工答复,以及逐接收人的通知结果和投递缺口。 - 为任务状态冲突提供结构化错误;项目、任务元数据支持版本条件读取和更新,降低并发操作覆盖较新状态的风险。 - 完善交付产物的 @mention 摘要通知、报告长度限制和评审上下文;创建项目时可通过 `--team` 邀请整个团队,并查看实际入群结果。 ### 多用户权限与模型授权 - 明确资源管理边界:Manager 和平台管理员负责资源管理;Team Leader 保留本团队任务协调、状态读取和自身就绪上报,不能直接管理其他 Worker 的配置与生命周期。 - 新增按用户、角色配置的模型授权及共享授权表,对 Worker 主模型、子 Agent 模型和 Coding CLI 模型的配置变更执行服务端校验。 - 新增工作区访问授权与限制,完善团队人类成员变更后的权限同步及已有团队房间的管理员授权。 - 人类用户列表不再返回初始密码,改进 Worker 凭据原子写入和人类用户并发更新时的凭据保留

  2. v1.2.5v1.2.5Oct 16, 202623 downloads

    # AgentTeams 1.2.5 Release Notes AgentTeams 1.2.5 重点增强编码工具委派、Worker 配置管理和 QwenPaw 运行时升级,默认搭配 Dashboard v1.3.1,并修复 MCP 注册、Matrix 消息协作及 Worker 恢复中的多项问题。 ## 新增与增强 ### 编码 CLI 委派与管理 - 新增 Qwen Code、OpenCode 执行适配,完善 Qoder CLI 接入及编码委派指引,让 Agent 可以调用外部编码工具完成任务。 - 提供 Worker 编码 CLI 的安装、状态检查、认证配置及执行参数管理能力;Qoder 支持通过 `QODER_PERSONAL_ACCESS_TOKEN` 环境变量注入 PAT。 - 新增编码执行记录与只读详情查询,支持查看关联任务、执行结果和日志;完善被阻止执行的记录展示及时间信息处理。 - 补齐 Qwen Code 的审批模式和执行时限参数,增强任务所有权检查及超时清理。 ### Worker 环境配置与 LLM 参数 - 新增 Worker 环境变量编辑与网关身份探测,便于配置 CLI 凭证及排查调用链路。 - 扩展 QwenPaw Worker 的 LLM 调优能力,支持重试次数、并发限制,以及流式响应的首段内容等待和输出间隔超时配置。 - 完善运行时代理接口的类型与权限检查,避免将某一运行时的专用操作错误地应用到其他 Worker。 ### QwenPaw 原地升级 - 新增升级前兼容性检查,提前发现运行环境和依赖不匹配。 - 为 QwenPaw Worker 提供原地运行时升级、升级状态查询和失败回滚能力。 - Manager 运行时与旧 CoPaw Worker 包解耦,减少运行时管理对旧包的依赖。 ### Dashboard v1.3.1 - 默认 Dashboard 更新至独立发布的 **v1.3.1**,补齐编码 CLI 管理界面,并支持 Qoder。 - 新增编码执行记录与详情页面,对齐执行记录读取接口。 - 补齐 LLM 参数调优入口,改善 Worker 运行时配置面板的加载与编辑体验。 - 修复 CLI 安装状态、认证状态和配置回显问题,并改善 Worker 详情弹窗布局。 ## 修复与稳定性 - **MCP 接入**:修正 API 地址的协议和端口解析;支持独立部署的 Higress Console 地址及登录会话;修复 QwenPaw Manager 的 `mcporter` 运行环境问题。 - **Matrix 消息协作**:区分空转回合与真正完成的回合,避免无产出时显示“已完成”;消费失败改为通用错误提示;尊重思考内容展示配置;完善群聊简写 @mention 和唤醒去重处理。 - **Worke

  3. v1.2.4v1.2.4Sep 20, 202625 downloads

    # AgentTeams 1.2.4 Release Notes AgentTeams 1.2.4 重点完善团队协作的任务状态、运行时管理和权限控制,升级 QwenPaw 运行时,并修复本地部署、升级及 Worker 启动中的多项问题。 ## 新增与增强 ### 更清晰的任务进度与执行记录 - TeamHarness 新增任务状态转换机制,记录状态历史、事件和进度,便于追踪任务执行过程。 - 持久化任务续执行状态,保留跨轮次执行所需的信息;任务提交时由代码发送完成通知,减少对模型主动汇报的依赖。 - 新增 Mermaid 工作流渲染和任务级检查 API,支持查看任务规格、状态转换及执行信息。 - Worker 心跳上报运行时任务状态,项目列表增加更新时间。 相关变更:[#1233](https://github.com/agentscope-ai/AgentTeams/pull/1233)、[#1183](https://github.com/agentscope-ai/AgentTeams/pull/1183)、[#1206](https://github.com/agentscope-ai/AgentTeams/pull/1206)、[#1230](https://github.com/agentscope-ai/AgentTeams/pull/1230)、[#1247](https://github.com/agentscope-ai/AgentTeams/pull/1247)。 ### 更完整的 QwenPaw 运行时管理 - Manager 与 Worker 升级至 QwenPaw 2.2.1,并向 Worker 传递模型能力配置。 - 新增子 Agent 独立模型配置,支持团队默认值和运行时热更新;配置修改受管理员权限控制。 - 新增 Worker 会话列表、详情和运行状态的只读代理接口,便于管理端查看会话执行情况。 - 扩展运行时配置、循环状态、频道配置与冲突检查、内置工具设置、技能运行状态及预加载策略等管理接口。 相关变更:[#1294](https://github.com/agentscope-ai/AgentTeams/pull/1294)、[#1295](https://github.com/agentscope-ai/AgentTeams/pull/1295)、[#1219](https://github.com/agentscope-ai/AgentTeams/pull/1219)、[#1255](https://github.com/agentscope-ai/AgentTeams/pull/1255)、[#1252](https://github.com/agentscope-ai/AgentTeams/pul

  4. v1.2.3v1.2.3Aug 22, 202635 downloads

    ## Highlights AgentTeams v1.2.3 makes long-running Project workflows visible and controllable from the Dashboard, promotes QwenPaw as the default local runtime, and improves installation and runtime reliability. ### What's New - **Human-visible Project workflows**: New Controller APIs and `agt` commands expose Project lists, workflow DAGs, task details, ready nodes, interrupts, artifacts, and spawned subagents. L2 humans can inspect only their authorized teams. ([#1169](https://github.com/agentscope-ai/AgentTeams/pull/1169)) - **Human intervention and lifecycle controls**: Projects can now be created, paused, resumed, replanned, and completed, and tasks can be cancelled through authenticated write APIs. Each intervention records audit metadata and a pre-change snapshot. ([#1172](https://github.com/agentscope-ai/AgentTeams/pull/1172), [#1191](https://github.com/agentscope-ai/AgentTeams/pull/1191)) - **Project history and Worker checkpoints**: New read APIs expose the Project intervention timeline and QwenPaw Worker checkpoint graph/status for diagnosis and recovery workflows. ([#1186](https://github.com/agentscope-ai/AgentTeams/pull/1186)) ### Dashboard Updates The installer no

  5. v1.2.2v1.2.2Aug 8, 202633 downloads

    ## Highlights **What's New** - **Manager-to-Worker custom Skill delivery**: A Manager can validate, upload, and assign a custom Worker Skill before updating `Worker.spec.skills`; QwenPaw Workers synchronize the assigned Skill into their native workspace and refresh plus enable it without a restart. ([c2d84b09](https://github.com/agentscope-ai/AgentTeams/commit/c2d84b09d3547b258decbb1d24b245c10a98c088), [#1153](https://github.com/agentscope-ai/AgentTeams/pull/1153)) **Bug Fixes** - **Worker custom Skill loading safeguards**: Grant Managers access to Worker storage prefixes, avoid recursive CR updates during reconciliation, keep local/default MinIO paths off unavailable STS, restore Manager MinIO access after Controller restarts, verify uploaded `SKILL.md` files, and restore Worker file-sync notifications. ([c2d84b09](https://github.com/agentscope-ai/AgentTeams/commit/c2d84b09d3547b258decbb1d24b245c10a98c088), [#1153](https://github.com/agentscope-ai/AgentTeams/pull/1153)) - **QwenPaw Manager custom Skill hot loading**: Continuously mirror workspace Skills into the native QwenPaw workspace, then refresh and enable additions or updates without restarting the Manager. ([c2d84b09](

Code frequency

additions and deletions
+137.9K-137.9KWeek of 2026-02-08: +7,365 linesWeek of 2026-02-08: -59 linesWeek of 2026-02-15: +17,959 linesWeek of 2026-02-15: -12,214 linesWeek of 2026-02-22: +6,572 linesWeek of 2026-02-22: -1,719 linesWeek of 2026-03-01: +18,880 linesWeek of 2026-03-01: -4,033 linesWeek of 2026-03-08: +20,912 linesWeek of 2026-03-08: -11,886 linesWeek of 2026-03-15: +18,954 linesWeek of 2026-03-15: -6,761 linesWeek of 2026-03-22: +14,523 linesWeek of 2026-03-22: -2,527 linesWeek of 2026-03-29: +15,810 linesWeek of 2026-03-29: -1,677 linesWeek of 2026-04-05: +519 linesWeek of 2026-04-05: -356 linesWeek of 2026-04-12: +37,256 linesWeek of 2026-04-12: -15,906 linesWeek of 2026-04-19: +29,229 linesWeek of 2026-04-19: -6,409 linesWeek of 2026-04-26: +2,854 linesWeek of 2026-04-26: -1,204 linesWeek of 2026-05-03: +1,567 linesWeek of 2026-05-03: -404 linesWeek of 2026-05-10: +717 linesWeek of 2026-05-10: -167 linesWeek of 2026-05-17: +19,146 linesWeek of 2026-05-17: -4,439 linesWeek of 2026-05-24: +777 linesWeek of 2026-05-24: -345 linesWeek of 2026-05-31: +3,794 linesWeek of 2026-05-31: -234 linesWeek of 2026-06-07: +1,438 linesWeek of 2026-06-07: -216 linesWeek of 2026-06-14: +2,658 linesWeek of 2026-06-14: -296 linesWeek of 2026-06-21: +13,541 linesWeek of 2026-06-21: -2,681 linesWeek of 2026-06-28: +14,068 linesWeek of 2026-06-28: -3,316 linesWeek of 2026-07-05: +50,490 linesWeek of 2026-07-05: -4,648 linesWeek of 2026-07-12: +8,486 linesWeek of 2026-07-12: -6,566 linesWeek of 2026-07-19: +99,956 linesWeek of 2026-07-19: -107,535 linesWeek of 2026-07-26: +13,200 linesWeek of 2026-07-26: -8,634 linesWeek of 2026-08-02: +6,368 linesWeek of 2026-08-02: -1,149 linesWeek of 2026-08-09: +32,156 linesWeek of 2026-08-09: -16,909 linesWeek of 2026-08-16: +2,691 linesWeek of 2026-08-16: -302 linesWeek of 2026-08-23: +0 linesWeek of 2026-08-23: -0 linesWeek of 2026-08-30: +5,533 linesWeek of 2026-08-30: -143 linesWeek of 2026-09-06: +5,254 linesWeek of 2026-09-06: -269 linesWeek of 2026-09-13: +35,460 linesWeek of 2026-09-13: -1,638 linesWeek of 2026-09-20: +4,335 linesWeek of 2026-09-20: -132 linesWeek of 2026-09-27: +845 linesWeek of 2026-09-27: -43 linesWeek of 2026-10-04: +18,830 linesWeek of 2026-10-04: -26,023 linesWeek of 2026-10-11: +17,547 linesWeek of 2026-10-11: -1,721 linesWeek of 2026-10-18: +7,068 linesWeek of 2026-10-18: -174 linesWeek of 2026-10-25: +47,182 linesWeek of 2026-10-25: -7,359 linesWeek of 2026-11-01: +137,907 linesWeek of 2026-11-01: -6,227 linesWeek of 2026-11-08: +15,602 linesWeek of 2026-11-08: -38,136 linesFeb 8, 2026Nov 8, 2026
+757.4K lines added, -304.5K removed over the last year.

Commits per week

last 52 weeks
1190Week of 2025-11-16: 0 commitsWeek of 2025-11-23: 0 commitsWeek of 2025-11-30: 0 commitsWeek of 2025-12-07: 0 commitsWeek of 2025-12-14: 0 commitsWeek of 2025-12-21: 0 commitsWeek of 2025-12-28: 0 commitsWeek of 2026-01-04: 0 commitsWeek of 2026-01-11: 0 commitsWeek of 2026-01-18: 0 commitsWeek of 2026-01-25: 0 commitsWeek of 2026-02-01: 0 commitsWeek of 2026-02-08: 3 commitsWeek of 2026-02-15: 23 commitsWeek of 2026-02-22: 27 commitsWeek of 2026-03-01: 111 commitsWeek of 2026-03-08: 119 commitsWeek of 2026-03-15: 105 commitsWeek of 2026-03-22: 38 commitsWeek of 2026-03-29: 38 commitsWeek of 2026-04-05: 14 commitsWeek of 2026-04-12: 26 commitsWeek of 2026-04-19: 55 commitsWeek of 2026-04-26: 12 commitsWeek of 2026-05-03: 17 commitsWeek of 2026-05-10: 9 commitsWeek of 2026-05-17: 35 commitsWeek of 2026-05-24: 11 commitsWeek of 2026-05-31: 12 commitsWeek of 2026-06-07: 7 commitsWeek of 2026-06-14: 6 commitsWeek of 2026-06-21: 4 commitsWeek of 2026-06-28: 8 commitsWeek of 2026-07-05: 13 commitsWeek of 2026-07-12: 22 commitsWeek of 2026-07-19: 5 commitsWeek of 2026-07-26: 19 commitsWeek of 2026-08-02: 11 commitsWeek of 2026-08-09: 11 commitsWeek of 2026-08-16: 9 commitsWeek of 2026-08-23: 0 commitsWeek of 2026-08-30: 5 commitsWeek of 2026-09-06: 8 commitsWeek of 2026-09-13: 45 commitsWeek of 2026-09-20: 16 commitsWeek of 2026-09-27: 8 commitsWeek of 2026-10-04: 22 commitsWeek of 2026-10-11: 18 commitsWeek of 2026-10-18: 13 commitsWeek of 2026-10-25: 26 commitsWeek of 2026-11-01: 66 commitsWeek of 2026-11-08: 44 commitsNov 16, 2025Nov 8, 2026
1K commits in the last 52 weeks.

When work happens

weekday and hour
SunMonTueWedThuFriSat036912151821Sun 0:00 — 10 commitsSun 1:00 — 7 commitsSun 2:00 — 0 commitsSun 3:00 — 2 commitsSun 4:00 — 2 commitsSun 5:00 — 0 commitsSun 6:00 — 1 commitsSun 7:00 — 1 commitsSun 8:00 — 2 commitsSun 9:00 — 5 commitsSun 10:00 — 9 commitsSun 11:00 — 9 commitsSun 12:00 — 3 commitsSun 13:00 — 3 commitsSun 14:00 — 3 commitsSun 15:00 — 9 commitsSun 16:00 — 11 commitsSun 17:00 — 8 commitsSun 18:00 — 3 commitsSun 19:00 — 5 commitsSun 20:00 — 8 commitsSun 21:00 — 8 commitsSun 22:00 — 13 commitsSun 23:00 — 2 commitsMon 0:00 — 15 commitsMon 1:00 — 0 commitsMon 2:00 — 0 commitsMon 3:00 — 1 commitsMon 4:00 — 1 commitsMon 5:00 — 0 commitsMon 6:00 — 1 commitsMon 7:00 — 1 commitsMon 8:00 — 4 commitsMon 9:00 — 2 commitsMon 10:00 — 7 commitsMon 11:00 — 6 commitsMon 12:00 — 1 commitsMon 13:00 — 4 commitsMon 14:00 — 9 commitsMon 15:00 — 12 commitsMon 16:00 — 8 commitsMon 17:00 — 5 commitsMon 18:00 — 0 commitsMon 19:00 — 6 commitsMon 20:00 — 5 commitsMon 21:00 — 6 commitsMon 22:00 — 12 commitsMon 23:00 — 8 commitsTue 0:00 — 11 commitsTue 1:00 — 2 commitsTue 2:00 — 0 commitsTue 3:00 — 0 commitsTue 4:00 — 0 commitsTue 5:00 — 0 commitsTue 6:00 — 0 commitsTue 7:00 — 0 commitsTue 8:00 — 3 commitsTue 9:00 — 10 commitsTue 10:00 — 9 commitsTue 11:00 — 9 commitsTue 12:00 — 5 commitsTue 13:00 — 2 commitsTue 14:00 — 11 commitsTue 15:00 — 12 commitsTue 16:00 — 6 commitsTue 17:00 — 5 commitsTue 18:00 — 7 commitsTue 19:00 — 4 commitsTue 20:00 — 8 commitsTue 21:00 — 4 commitsTue 22:00 — 11 commitsTue 23:00 — 6 commitsWed 0:00 — 3 commitsWed 1:00 — 4 commitsWed 2:00 — 2 commitsWed 3:00 — 0 commitsWed 4:00 — 0 commitsWed 5:00 — 0 commitsWed 6:00 — 0 commitsWed 7:00 — 0 commitsWed 8:00 — 1 commitsWed 9:00 — 5 commitsWed 10:00 — 6 commitsWed 11:00 — 17 commitsWed 12:00 — 3 commitsWed 13:00 — 13 commitsWed 14:00 — 16 commitsWed 15:00 — 10 commitsWed 16:00 — 21 commitsWed 17:00 — 4 commitsWed 18:00 — 26 commitsWed 19:00 — 3 commitsWed 20:00 — 5 commitsWed 21:00 — 10 commitsWed 22:00 — 9 commitsWed 23:00 — 10 commitsThu 0:00 — 7 commitsThu 1:00 — 5 commitsThu 2:00 — 1 commitsThu 3:00 — 2 commitsThu 4:00 — 0 commitsThu 5:00 — 0 commitsThu 6:00 — 1 commitsThu 7:00 — 0 commitsThu 8:00 — 2 commitsThu 9:00 — 3 commitsThu 10:00 — 5 commitsThu 11:00 — 13 commitsThu 12:00 — 8 commitsThu 13:00 — 9 commitsThu 14:00 — 16 commitsThu 15:00 — 6 commitsThu 16:00 — 10 commitsThu 17:00 — 15 commitsThu 18:00 — 5 commitsThu 19:00 — 7 commitsThu 20:00 — 6 commitsThu 21:00 — 15 commitsThu 22:00 — 6 commitsThu 23:00 — 11 commitsFri 0:00 — 10 commitsFri 1:00 — 7 commitsFri 2:00 — 1 commitsFri 3:00 — 5 commitsFri 4:00 — 3 commitsFri 5:00 — 0 commitsFri 6:00 — 0 commitsFri 7:00 — 0 commitsFri 8:00 — 3 commitsFri 9:00 — 4 commitsFri 10:00 — 6 commitsFri 11:00 — 11 commitsFri 12:00 — 5 commitsFri 13:00 — 5 commitsFri 14:00 — 11 commitsFri 15:00 — 13 commitsFri 16:00 — 13 commitsFri 17:00 — 16 commitsFri 18:00 — 10 commitsFri 19:00 — 8 commitsFri 20:00 — 10 commitsFri 21:00 — 12 commitsFri 22:00 — 10 commitsFri 23:00 — 22 commitsSat 0:00 — 11 commitsSat 1:00 — 4 commitsSat 2:00 — 4 commitsSat 3:00 — 0 commitsSat 4:00 — 0 commitsSat 5:00 — 0 commitsSat 6:00 — 0 commitsSat 7:00 — 4 commitsSat 8:00 — 3 commitsSat 9:00 — 6 commitsSat 10:00 — 8 commitsSat 11:00 — 5 commitsSat 12:00 — 5 commitsSat 13:00 — 13 commitsSat 14:00 — 6 commitsSat 15:00 — 5 commitsSat 16:00 — 17 commitsSat 17:00 — 13 commitsSat 18:00 — 7 commitsSat 19:00 — 5 commitsSat 20:00 — 15 commitsSat 21:00 — 9 commitsSat 22:00 — 16 commitsSat 23:00 — 18 commits
Commit volume by weekday and hour (UTC). Larger dots mean more commits.
DateListRankStars gained
Mar 6, 2026daily#22+168
  • openclaw/openclaw

    The AI that really does things. Any OS. Any Platform. The lobster way. 🦞

    394.9K stars · TypeScript

  • avelino/awesome-go

    A curated list of awesome Go frameworks, libraries and software

    195K stars · Go

  • Graphify-Labs/graphify

    Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.

    151.2K stars · Python

  • stablyai/orca

    Orca lets you run a fleet of bots with your own subscription (e.g. Claude Code, Grok, Agy, Codex). Available on desktop, mobile and remote runtime.

    131.1K stars · TypeScript

  • kubernetes/kubernetes

    Production-Grade Container Scheduling and Management

    129.3K stars · Go

  • JuliusBrussee/caveman

    why use many token when few token do trick. Skill + local proxy for coding agents: shorter answers, smaller tool output, every original recoverable.

    121.8K stars · Go