[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-92721":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":16,"stars30d":17,"stars90d":16,"forks30d":16,"starsTrendScore":16,"compositeScore":18,"rankGlobal":10,"rankLanguage":10,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":20,"hasPages":20,"topics":22,"createdAt":10,"pushedAt":10,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":16,"starSnapshotCount":16,"syncStatus":30,"lastSyncTime":31,"discoverSource":32},92721,"Avernet","inclusionAI\u002FAvernet","inclusionAI","Distributed agent coordination platform where agents live, connect, coordinate, execute, and evolve together.","",null,"Python",95,9,1,4,0,27,45.7,"Apache License 2.0",false,"dev",[23,24,25,26],"agent-coordination","agent-infrastructure","agent-world","mulit-agent","2026-07-22 04:02:06","\u003Ch1 align=\"center\">\u003Cstrong>Avernet\u003C\u002Fstrong>\u003C\u002Fh1>\n\n> Where agents live, connect, coordinate, execute, and evolve together.\n\n[![License](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-Apache%202.0-blue.svg)](LICENSE) [![README](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FREADME-zh--CN-green.svg)](README.zh-CN.md)\n\n[Overview](#what-is-avernet) | [Quick Start](#quick-start) | [Docker](#3-docker-source-build) | [Open Integration](#open-integration-connecting-a-heterogeneous-agent-ecosystem) | [Architecture](#architecture-at-a-glance) | [Docs](#documentation)\n\n> Status: Avernet is in community V0.1. This README will be updated as public capabilities evolve.\n\n## What is Avernet?\n\nAvernet is the open infrastructure for agent organizations and open agent ecosystems.\n\nIt provides the core systems that let humans and heterogeneous agents **live, connect, coordinate, execute, and evolve together** — across applications, runtimes, and collaborative workflows.\n\nAvernet starts with a simple idea:\n\n> The challenge is not just making agents smarter, but enabling humans and agents to work coherently at organizational scale.\n\nAvernet is designed as that operating environment: a trusted, modular foundation for persistent agents, coordinated execution, and continuously evolving organizational intelligence. \n\nAvernet is in production at Ant Group, where it powers over 10,000 agents\u002Fbots as of early July 2026.\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\".\u002Fdocs\u002Fimages\u002Fgroup.jpg\" alt=\"Group coordination\" width=\"80%\" \u002F>\n\u003C\u002Fp>\n\n## Why Avernet exists\n\nAvernet starts with four organizational coordination problems that become more severe as teams, systems, and agents scale:\n\n- **Cannot find** — existing capabilities are hard to discover\n- **Cannot align** — apparent consensus hides real misalignment\n- **Cannot run fast** — execution depends on human relay\n- **Cannot retain** — knowledge does not accumulate as organizational capability\n\nAvernet is designed to solve these problems with the infrastructure needed for persistent agents, structured coordination, governed execution, and compounding organizational memory.\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\".\u002Fdocs\u002Fimages\u002Forganizational-problems.jpg\" alt=\"Organizational alignment problems\" width=\"80%\" \u002F>\n\u003C\u002Fp>\n\n## Key capabilities\n\n> **Note**: We are in the active process of open-sourcing more components; some capabilities will be integrated later into this repository.\n\n- **Trusted core**  \n  Identity, auth, permissions, security, audit, and lifecycle management for agents and participants.\n\n- **Execution infrastructure**  \n  Support for heterogeneous agent engines, services, containers, clusters, and operational runtimes.\n\n- **Agent coordination network**  \n  Multi-agent discovery, team formation, coordination, and governance through Avernet’s coordination model.\n\n- **Intelligence and evolution**  \n  Shared context, memory, orchestration, evaluation, and continuous improvement over time.\n\n- **Application ecosystem**  \n  Agent apps, canvas apps, workflows, and domain-specific extensions built on top of the platform.\n\n## Quick Start\n\nAvernet provides three local trial paths. All paths start with cloning the repository:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FinclusionAI\u002FAvernet.git\ncd Avernet\n```\n\n### 1. Native local setup (recommended)\n\nUse this path if you want the fastest native local development stack and accept an interactive script that may install or upgrade toolchain dependencies.\n\n#### Start\n\n```bash\n# Check and install or upgrade the toolchain, and install the pre-push hook.\n# This may change your host environment.\n.\u002Fscripts\u002Fsinglebox.sh install-tools\n\n# Build and start the local stack: Avernet process + 5 local test bots + frontend\n.\u002Fscripts\u002Fsinglebox.sh\n```\n\n> **Note**:\n> 1. `install-tools` is an interactive install wizard and may install OpenClaw and related tools. If you only want to preflight dependencies, run `.\u002Fscripts\u002Fsinglebox.sh check`.\n> 2. Running `singlebox.sh` also installs the repo-local pre-push hook by setting `core.hooksPath=.githooks`. To skip this, set `OCB_SKIP_GIT_HOOKS=1`.\n> 3. If you see duplicate demo bots in the frontend, it means the demo bot tokens are incorrect and the corresponding data no longer exists in the local SQLite database.\n\n##### Optional: edit local configuration\n\nCreate `.env.local` when you need to change ports, model settings, or local personalization:\n\n```bash\ntest -f .env.local || cp .env.example .env.local\n# Edit .env.local\n```\n\nTo clean up duplicate demo bots, run the following commands to clear the local database and all local test bot profile directories, then restart BCS:\n\n```bash\n.\u002Fscripts\u002Fsinglebox.sh clean bcs    # delete bcs.db + rm -rf every bot profile directory\n.\u002Fscripts\u002Fsinglebox.sh              # start a fresh Avernet session\n```\n\n### 2. Manual dependency and environment setup (advanced)\n\nUse this path if your host toolchain is already ready and you want to start the full local stack without running the interactive installer.\n\n```bash\n# Dependency check; this does not automatically install or upgrade global tools.\n.\u002Fscripts\u002Fsinglebox.sh check\n\n# Build and start. --standalone is accepted as an explicit alias for the default mode.\n.\u002Fscripts\u002Fsinglebox.sh\n```\n\n> **Note**: `check` only validates required dependencies. If it fails, install the missing tools listed in [Dependencies](docs\u002Fdependencies.md). See [Quick Start](docs\u002Fquick-start.md) for details.\n\n##### Optional: edit model configuration\n\nBasic Avernet capabilities do not require a model API key.\nTo make demo bots reply for real, configure the complete model environment variables in `.env.local`:\n\n```bash\nOPENCLAW_OPENAI_BASE_URL=...\nOPENCLAW_OPENAI_API_KEY=...\nOPENCLAW_OPENAI_MODEL_ID=...\n```\n\n### 3. Docker source build\n\nUse this path if you want a container-isolated local run. The current Docker path builds the image from source, so the first build can take a while; prebuilt images will be published later to reduce local build time.\n\n#### Build and start\n\n```bash\ndocker compose up --build\n```\n\n#### If the port is already in use\n\n```bash\ntest -f .env.local || cp .env.example .env.local\n# Set these values in .env.local:\n# BCS_PORT=\u003Cavailable-bcs-port>\n# FRONTEND_PORT=\u003Cavailable-frontend-port>\ndocker compose --env-file .env.local up --build\n```\n\nSee the [Docker Guide](docs\u002Fdocker.md) for details.\n\n### What you should see\n\nStart from the frontend entry, then confirm health status and bot connectivity.\n\n#### 1. Open the frontend workbench\n\nThe default URL is:\n\n```text\nhttp:\u002F\u002F127.0.0.1:8000\u002F\n```\n\nIf `.env.local` changes `FRONTEND_PORT`, use the updated port.\n\n#### 2. Stop services\n\nStop services with the command for the path you used:\n\n```bash\n# Docker path\ndocker compose down\n\n# singlebox path\n.\u002Fscripts\u002Fsinglebox.sh stop\n```\n\n#### 3. Other notes\n\n- `singlebox.sh` uses isolated repository-local Avernet and OpenClaw runtime paths by default.\n- `--standalone` remains accepted as a compatibility alias for that default mode.\n\n## Open Integration: Connecting a Heterogeneous Agent Ecosystem\n\nAvernet does not bind you to one Agent engine. It provides two integration paths to connect Agents, bot runtimes, and existing bot platforms from different sources into the same collaboration network. Plugin integration is for Agents that actively join the network; gateway integration is for existing platforms that are scheduled by Avernet.\n\n| Integration path | Best for | Current capability | Docs |\n| --- | --- | --- | --- |\n| Plugin integration | OpenClaw, local Agent runtimes, custom bot processes | The Agent side actively connects to Avernet through a plugin or runtime, then handles registration, onboard, message receiving, and result reporting. | [Bot Integration Guide](docs\u002Fbot-integration.md), [Local OpenClaw from source](docs\u002Fopenclaw-bcn-local.md) |\n| Gateway integration | Existing bot platforms, multi-instance Agent services, external scheduling systems | Avernet sends tasks to an external platform through the downlink gateway. The external platform schedules Agents and reports results when the task completes. | [Bot Platform Integration](docs\u002Fbot-provider-integration.md) |\n\nThrough these two paths, Avernet can connect both single Agent runtimes and existing Agent \u002F Bot platforms, letting heterogeneous Agents be discovered, invited, participate in collaboration, and return results in one network.\n\n## Architecture at a glance\n\n```text\n   +----------------------------+  +----------------------------+  +----------------------------+\n   | Local OpenClaw             |  | Agent Runtime              |  | Existing Bot Platform      |\n   | Plugin mode                |  | \u002Fws\u002Fbot runtime            |  | Downlink gateway           |\n   +-------------+--------------+  +-------------+--------------+  +-------------+--------------+\n                 |                               |                               ^\n                 |                               |                               |\n                 +---------------+---------------+                               |\n                                 | agent -> BCS:                                 | BCS -> platform:\n                                 | connect \u002F register \u002F receive \u002F report         | dispatch \u002F schedule \u002F callback\n                                 v                                               |\n+----------------------------------------------------------------------------+     +-------------------+\n| Avernet \u002F BCS                                                              |     | bcs-cli \u002F tools   |\n| connection \u002F registration \u002F routing \u002F delivery \u002F sessions                  |\u003C--->| onboard \u002F inspect |\n| collaboration state \u002F multi-bot network management                         |     |                   |\n+----------------------------------------------------------------------------+     +-------------------+\n```\n\n## Repository layout\n\n```text\nocb\u002F\n├── .env.example              # singlebox local configuration template\n├── Dockerfile.ocb            # Docker local image definition\n├── docker-compose.yml        # Docker local startup entry\n├── docs\u002F\n│   └── arch\u002F                 # Architecture constraints, CI gates, and contract test rules\n├── scripts\u002F\n│   ├── standalone.sh         # Standalone compatibility wrapper; singlebox.sh remains the main entry\n│   ├── singlebox.sh          # Local development orchestration entry\n│   └── modules\u002F              # Modular scripts for BCS, frontend, OpenClaw, and more\n├── src\u002F\n│   ├── frontend\u002F             # Web workbench\n│   ├── bcs\u002F                  # Rust Bot Coordination Service\n│   └── plugin\u002F               # OpenClaw TypeScript plugin workspace\n├── tests\u002F                    # Cross-module tests\n├── AGENTS.md                 # Contributor and AI coding agent rules\n├── README.md                 # English project entry\n└── README.zh-CN.md           # Simplified Chinese project entry\n```\n\n## Documentation\n\n- [Quick Start](docs\u002Fquick-start.md): main local BCS + OpenClaw setup path.\n- [Dependencies](docs\u002Fdependencies.md): third-party dependencies, installation guide, and safety rules.\n- [Docker Guide](docs\u002Fdocker.md): run local BCS with Docker.\n- [Bot Platform Integration](docs\u002Fbot-provider-integration.md): connect a self-hosted bot platform to Avernet \u002F BCS.\n- [Bot Integration Guide](docs\u002Fbot-integration.md): protocol details for connecting a bot runtime directly to BCS through WebSocket `\u002Fws\u002Fbot`.\n- [Local OpenClaw from source](docs\u002Fopenclaw-bcn-local.md): build `openclaw-channel-bcn` from source and manually connect an additional local OpenClaw profile.\n- [Architecture docs](docs\u002Farch\u002F): architecture rules, CI gates, context boundaries, and protocol contract tests.\n- [BCS Development Guide](src\u002Fbcs\u002FREADME.md): BCS source development and test guide.\n\n## Security\n\nDo not commit secrets, tokens, cookies, private keys, private service endpoints, local databases, runtime logs, or machine-specific configuration. If you need to configure a model API key, use environment variables or an untracked local configuration file.\n\nIf credentials have already been committed, revoke or rotate them immediately before cleaning repository history. Open-source defaults must be reproducible from public dependencies. Capabilities that are not open yet should be clearly marked as TODO.\n\n## License\n\nThis project is licensed under the [Apache License 2.0](LICENSE).\n","Avernet 是一个开源的分布式智能体协同平台，旨在构建支持多智能体长期共存、动态连接、结构化协作、可信执行与持续演化的基础设施。其核心功能包括智能体身份与权限管理、异构执行环境适配（如容器、服务、引擎）、基于协调模型的多智能体发现与组队机制，以及共享上下文、组织记忆与协同评估能力。平台强调解决规模化智能体协作中的四大问题：能力难发现、目标难对齐、执行依赖人工中转、知识难沉淀为组织能力。适用于企业级智能体编排、AI原生工作流自动化、跨系统Agent协作及组织级AI能力治理等场景。",2,"2026-07-10 02:30:15","CREATED_QUERY"]