[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-95119":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":9,"totalLinesOfCode":9,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":9,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":17,"stars30d":17,"stars90d":16,"forks30d":16,"starsTrendScore":17,"compositeScore":18,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":19,"fork":19,"defaultBranch":20,"hasWiki":19,"hasPages":19,"topics":9,"createdAt":9,"pushedAt":9,"updatedAt":21,"readmeContent":22,"aiSummary":23,"trendingCount":16,"starSnapshotCount":16,"syncStatus":14,"lastSyncTime":24,"discoverSource":25},95119,"autonomous-os","autonomous-ai\u002Fautonomous-os","autonomous-ai","The open-source operating system for robots — install it and your robot comes alive",null,"https:\u002F\u002Fgithub.com\u002Fautonomous-ai\u002Fautonomous-os","Python",139,16,2,14,0,8,45.49,false,"main","2026-08-24 04:01:23","## Autonomous OS: The \"Android\" for Robots\n\nRobots have been around for years but have never been autonomous — someone has to drive them with a remote, and they've stopped at scripted demos. Autonomous OS brings autonomy to robots: install it on your robot and it comes alive.\n\n- **Your robot thinks.** Everything it sees and hears goes to an agentic reasoning engine running on the robot itself — [Hermes](runtimes\u002Fhermes\u002F), [Claude Code](runtimes\u002Fclaudecode\u002F), or [whichever you choose](runtimes\u002F) — that decides what to do next.\n- **Your robot acts.** It [guards the house](skills\u002Fguard\u002F), [knows your face](skills\u002Fface-enroll\u002F), [follows you as you move](skills\u002Fservo-tracking\u002F), [reads the mood on your face](skills\u002Fuser-emotion-detection\u002F), [sets the light](skills\u002Fscene\u002F) — and does the desk work too: [Gmail, GitHub](skills\u002Fconnectors\u002F), [your Mac](skills\u002Fcomputer-use\u002F). Each one is a [skill](skills\u002F) — install more from the Skill Store, or write your own.\n- **Your robot grows.** It has a built-in learning loop. It creates skills from experience, sharpens them as it uses them, keeps what it learns, searches its own past conversations, and builds a deeper picture of you with every session.\n\nAutonomous OS is a fully customizable operating system for robots. Every component is swappable — [engine](runtimes\u002F), [model](docs\u002Fhosted.md), [voice](hal\u002Fdrivers\u002Fvoice\u002F), [skills](skills\u002F), [board](hal\u002Fboard\u002Fboards.json). Your robot declares what it has in a `ROBOT.md`, and the OS mounts exactly that. When a better one ships, your robot gets it the same day — and gets better without new hardware.\n\n## Quick start\n\nThe simplest way in is a robot we have already tested it on. What each of them can do: [robot comparison](docs\u002Frobot-comparison.md).\n\n### Autonomous Lamp\n\n[Lamp](https:\u002F\u002Fwww.autonomous.ai\u002Flamp) is the robot that shows the whole OS — it sees, hears, speaks, moves, and ships with Autonomous OS on it.\n\n\u003Cimg src=\"robots\u002Flamp\u002Fimages\u002Flamp-hero.webp\" alt=\"Autonomous Lamp on a desk, ring lit\">\n\n1. **Add it.** In the Autonomous app ([iOS](https:\u002F\u002Fapps.apple.com\u002Fapp\u002Fid6744885683) | [Android](https:\u002F\u002Fplay.google.com\u002Fstore\u002Fapps\u002Fdetails?id=ai.autonomous.connect.wifi)), tap **Add robot → Lamp**.\n2. **Set up Wi-Fi.** Pick your network in the app; it joins the robot's hotspot and hands over the keys and pairing.\n3. **Interact with Lamp.** Say something, it turns to look at you, the ring lights up, and it answers.\n4. **Install a skill** from the Skill Store — one tap, live on the next conversation.\n5. **Build your own skill.** Type what you want it to do in the app and it writes the skill.\n6. **Give it a character.** Edit [`SOUL.md`](robots\u002Flamp\u002FSOUL.md) and it is someone else on the next turn.\n\n### Reachy Mini\n\n[Reachy Mini](https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Freachy_mini) is Hugging Face's desk robot, running our OS beside its own stack.\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F2f0aaafb-287c-488e-a3b1-a82f0ad9e776\n\n1. **SSH in** — `ssh pollen@reachy-mini.local`.\n2. **Run one command.** Nothing is flashed; the Reachy daemon keeps the motors.\n   ```bash\n   curl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002Fautonomous-ai\u002Fautonomous-os\u002Fmain\u002Frobots\u002Freachy-mini\u002Finstall.sh | sudo bash\n   ```\n3. **Add it.** In the app, tap **Add robot → Reachy Mini** and give it `reachy-mini.local`.\n4. **Interact with it.** Say something — the head tilts, the antennas lift, and it answers.\n5. **Install a skill** from the Skill Store, or type what you want it to do and it writes one.\n6. **Give it a character.** Edit `\u002Fopt\u002Fdevices\u002Freachy-mini\u002FSOUL.md`. Everything else, including how to undo the install: [`devices\u002Freachy-mini\u002FREADME.md`](robots\u002Freachy-mini\u002FREADME.md).\n7. **Put it next to a Lamp.** Each one hears the other's answer as its next input, so the two of them will hold a conversation until you stop them.\n\n### Autonomous Intern\n\n[Intern](https:\u002F\u002Fwww.autonomous.ai\u002Fintern) is the always-on desk agent: mic, speaker, LED ring.\n\n\u003Cimg src=\"robots\u002Fintern-v2\u002Fimages\u002Fintern-hero.webp\" alt=\"Autonomous Intern on a desk beside a laptop, tip glowing blue\">\n\n1. **Add it.** In the app, tap **Add robot → Intern**.\n2. **Set up Wi-Fi.** Same flow as Lamp: pick your network and it handles the keys and pairing.\n3. **Interact with it.** Say something and it answers; the ring shows what it is doing.\n4. **Install a skill** from the Skill Store.\n5. **Build your own skill.** Type what you want in the app; it is live on the next conversation.\n6. **Give it a character.** Edit `\u002Fopt\u002Fdevices\u002Fintern-v2\u002FSOUL.md`.\n\n## Bring your own robot\n\nAutonomous OS runs on any robot you can describe in four markdown files.\n\n- **`ROBOT.md`** — the body: the board and the hardware it has.\n- **`SOUL.md`** — the self: who it is and how it talks.\n- **`SAFETY.md`** — the bounds: how fast, how bright, how late.\n- **`SKILL.md`** — the hands: one thing it can do.\n\nFollow **[the full guide](docs\u002Fbring-your-own-robot.md)**.\n\n## Platform architecture\n\nAutonomous OS is a software stack. Each layer uses only the layer below it, so any layer can be replaced without touching the others. Every layer is a folder in this repo.\n\n![Autonomous OS stack, top down: apps, skills, the agentic runtime, the Go system services, the realtime voice agent, the capabilities a robot declares, the safety gate, drivers, boards, the vendor Linux kernel, and the bodies — one colour per layer, and the rows you can extend yourself drawn dashed](docs\u002Farchitecture\u002Fautonomous-stack.png)\n\n### [Apps](system\u002Fweb\u002F)\n\nWhat a person touches. The Autonomous app adds a robot, sets up Wi-Fi, installs skills from the Skill Store and switches brains; the robot also serves its own setup and monitor UI from `system\u002Fweb\u002F`. Both talk to os-server on :5000.\n\n### [Skills](skills\u002F)\n\nOne folder per behavior, one `SKILL.md` inside: markdown the agent reads. A skill acts by writing `[HW:\u002Fpath:{json}]` markers in its reply, so it never touches a servo bus or a GPIO pin. Each skill declares the capabilities it needs and installs on every robot that has them.\n\n### [Agentic runtime](runtimes\u002F)\n\nThe engine that thinks. Six of them — Hermes, OpenClaw, PicoClaw, Codex, Claude Code, OpenCode — behind one 76-method `AgentGateway`. It reads the robot's `SOUL.md` and its installed skills. Switch live from the web UI; persona, memory and connectors move with it.\n\n### [System services](system\u002F)\n\nThe Go daemon `os-server` on :5000, one package per box in the figure. `intent` answers fixed commands from a local table with no model; `server` strips `[HW:…]` markers out of a reply and POSTs them to HAL before the words are spoken; `agent` switches engines; `bootstrap` is OTA, its own binary.\n\n### [Realtime voice](hal\u002Frealtime\u002F)\n\nGemini Live, OpenAI Realtime or Qwen, hosted inside HAL and running beside the main path. A spoken turn lands here first: it answers directly, or hands the turn up to the engine.\n\n### [Capabilities](robots\u002Fcontract\u002Fcapabilities.md)\n\nThe 13 names a robot may declare — audio, vision, sensing, presence, motion, light, display, expression, lifelike, media, connectivity, companion, system. Ten mount [HTTP routes](hal\u002Froutes\u002F) on :5001 (111 endpoints, live Swagger at `\u002Fapi\u002Fhardware\u002Fdocs`); `presence` and `lifelike` are loops with no route, `companion` lives in os-server. HAL mounts only what `ROBOT.md` declares and fails loud on a missing required driver.\n\n### [Safety gate](hal\u002Fsafety\u002F)\n\nA pure function of `SAFETY.md`, below the engine and in every request path: brightness, quiet hours, explicit-move speed. No model in the loop — the same clamp whoever asked. What it does not cover yet: [`docs\u002Fsafety.md`](docs\u002Fsafety.md).\n\n### [Drivers](hal\u002Fdrivers\u002F)\n\nOne folder per subsystem: motors, rgb, camera, voice, display, sensing, tracking, and the media handover a third-party daemon needs. New hardware is one class and one factory line.\n\n### [Boards](hal\u002Fboard\u002Fboards.json)\n\nOne JSON entry per board, matched against `\u002Fproc\u002Fdevice-tree\u002Fmodel`. Raspberry Pi 4, Pi 5, CM4 and OrangePi 4 Pro today. A new board is an entry, not a code change.\n\n### Linux\n\nThe vendor kernel — Raspberry Pi OS, OrangePi Debian, or the robot's own image. We do not ship one, and nothing above the drivers has a real-time deadline: position control closes in the servo firmware, or in the robot's own daemon.\n\n### [Bodies](robots\u002F)\n\nFour markdown files and a driver per robot. Declarations, not forks — a body is a PR.\n\nLong form: [architecture](docs\u002Farchitecture\u002Foverview.md) · [HAL](docs\u002Farchitecture\u002Fhal.md) · [device spec](robots\u002Fcontract\u002FROBOT-SPEC.md) · [capabilities](robots\u002Fcontract\u002Fcapabilities.md) · [safety](docs\u002Fsafety.md) · [developer guide](docs\u002Fdeveloper-guide.md).\n\n## Contribute\n\nThe easiest way in is a skill: one markdown file, no Go, no hardware, and it lands on every robot that has the parts. PRs welcome, vibe-coded ones included. Questions, half-built ports and show-and-tell go in [Discussions](https:\u002F\u002Fgithub.com\u002Fautonomous-ai\u002Fautonomous-os\u002Fdiscussions); gaps we would love help with are labelled [`claim-me`](https:\u002F\u002Fgithub.com\u002Fautonomous-ai\u002Fautonomous-os\u002Fissues?q=is%3Aissue+is%3Aopen+label%3Aclaim-me) — comment to take one.\n\n| You want to… | You write… | Start from |\n|---|---|---|\n| Teach every robot something new | `skills\u002F\u003Cname>\u002FSKILL.md` (+ `skill.json` if it needs hardware) | [`skills\u002Fguard\u002F`](skills\u002Fguard\u002F) · [`skill-creator`](skills\u002Fskill-creator\u002F) |\n| Run Autonomous on your robot | `robots\u002F\u003Cid>\u002FROBOT.md` + `SAFETY.md` + `SOUL.md` | [`robots\u002Freachy-mini\u002F`](robots\u002Freachy-mini\u002F) — a third-party port, end to end |\n| Support new hardware | a class in `hal\u002Fdrivers\u002F\u003Csubsystem>\u002F` + one factory line | [`reachy_service.py`](hal\u002Fdrivers\u002Fmotors\u002Freachy_service.py) |\n| Support a new board | one entry in `hal\u002Fboard\u002Fboards.json` | [`boards.json`](hal\u002Fboard\u002Fboards.json) |\n| Add a brain | an `AgentGateway` implementation in `runtimes\u002F\u003Cname>\u002F` | [`adding-agent-runtime.md`](docs\u002Fagentic\u002Fadding-agent-runtime.md) |\n\nSeven more paths — apps, chat bridges, perception models, voices, safety bounds, CTS probes — and the norms: [`CONTRIBUTING.md`](CONTRIBUTING.md). One rule worth knowing up front: [`robots\u002Fcontract\u002F`](robots\u002Fcontract\u002F) is the interface everyone builds on, so open an issue before you change it.\n\nBuild locally:\n\n```bash\nmake os-build && make os-test          # Go daemon, cross-compiled to linux\u002Farm64\n(cd hal && uv sync) && make hal-dev    # HAL on :5001 with reload\nmake web-install && make web-dev       # setup + monitor UI\nmake cts                               # is this a valid Autonomous device?\n```\n\n## License\n\nEverything outside `hal\u002F` is Apache-2.0. `hal\u002F` is GPL-3.0, kept that way by choice so the tree has one license per top-level folder; a driver you commit there is GPL, so a closed vendor SDK wraps out of process.\n\nA robot running this carries other people's work: Pollen's [`reachy_mini`](https:\u002F\u002Fgithub.com\u002Fpollen-robotics\u002Freachy_mini) SDK, [YOLOv8](https:\u002F\u002Fgithub.com\u002Fultralytics\u002Fultralytics) for tracking (AGPL-3.0 — read it before you ship), [TEN-VAD](https:\u002F\u002Fgithub.com\u002FTEN-framework\u002Ften-vad) and [Silero](https:\u002F\u002Fgithub.com\u002Fsnakers4\u002Fsilero-vad) for hearing, [LeRobot](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Flerobot) and the [LeLamp Runtime](https:\u002F\u002Fgithub.com\u002Fhumancomputerlab\u002Flelamp_runtime) under the motion code, and the brains we install but do not ship. All of it, including what we copied verbatim: [`CREDITS.md`](CREDITS.md). Security issues: [`SECURITY.md`](SECURITY.md).\n","Autonomous OS 是一个面向机器人的开源操作系统，旨在赋予机器人本地化自主决策与持续学习能力。其核心功能包括：集成轻量级大模型推理引擎（如 Hermes、Claude Code）实现端侧智能决策；模块化技能系统（Skill Store），支持人脸识别、情绪感知、环境交互、办公自动化等即插即用能力；基于 ROBOT.md 的硬件抽象层，实现跨平台适配与热更新。系统强调“机器人自生长”，通过对话记忆、技能演化与经验沉淀构建个性化行为模型。适用于教育机器人、家庭服务机器人、桌面实验平台等需长期人机协同与渐进式智能演化的场景。","2026-08-21 02:30:13","trending"]