δΈζ | ζ₯ζ¬θͺ | νκ΅μ΄ | PortuguΓͺs | TiαΊΏng Viα»t | FranΓ§ais | Italiano | Bahasa Indonesia | Malay | English
Rhizome is a community-maintained hard fork of PicoClaw. It is written entirely in Go and aims to put a personal AI agent on every machine you already own β aging laptops, old phones, single-board computers, free-tier VMs.
Rhizome is a personal AI assistant inspired by NanoBot. It adds a Go-native P2P mesh, workspace sync, and an agent gateway on top of the original PicoClaw idea.
Single Go binary, no runtime dependencies β runs natively on Linux, Windows, macOS, FreeBSD/NetBSD, and Android, including 32-bit x86, ARMv7, and Android 5.0+ (API 21) targets. See the Hardware Compatibility List for verified boards and the current two-tier resource requirements.
Caution
Security Notice
- NO CRYPTO: Rhizome has not issued any official tokens or cryptocurrency. All claims on
pump.funor other trading platforms are scams. - CANONICAL SOURCE: The canonical source and release location is https://github.com/stpinkie/rhizome; releases are published under GitHub Releases. Beware of third-party domains claiming to be official.
- BEWARE: Many
.ai/.org/.com/.net/...domains have been registered by third parties. Do not trust them. - NOTE: Rhizome is in early rapid development. There may be unresolved security issues. Do not deploy to production before v1.0.
- NOTE: The full
rhizomebinary is ~74 MB stripped and the daemon uses ~60 MB private memory. Resource optimization is planned after feature stabilization.
2026-09-08 π v0.7.1 β Stabilization, Security & Browser Automation hardens the build (CI/test fixes, lint at zero issues), adds shared secret redaction in logs and tool output, prompt-injection screening for tool calls, and ships opt-in browser automation: eight browser_* agent tools across 12 pluggable backends (local agent-browser/system Chrome, custom CDP, Browserbase/Browserless/Kernel/Steel/Hyperbrowser/TinyFish clouds, Cloudflare REST), plus a /browser admin page and a Patchright-based stealth path via MCP. See the v0.7.1 release notes.
2026-09-07 π v0.7.0 β Swarm Mode adds named groups of trusted mesh peers with signed-envelope coordination over /rhizome/swarm/1.0.0: presence and roster gossip, an offer/claim distributed work queue, deterministic coordinator election with shared state synced through the workspace, goal orchestration (rhizome swarm run), a Swarms panel on the Network dashboard, and an optional GossipSub transport. See the v0.7.0 release notes.
2026-09-08 π v0.6.2 β Mesh resilient routing adds peer performance scoring (~/.rhizome/mesh-peer-scores.json) to Mesh.PickPeer, remote task retry/failover across capable peers, and a validated inline peer-list editor in the web console Mesh config. See the v0.6.2 release notes.
2026-09-07 π v0.6.1 β Mesh operations follow-through adds a persisted task store (~/.rhizome/mesh-tasks.jsonl) so remote tasks survive daemon restarts, and a live GET /api/network/tasks/events Server-Sent Event stream that drives the Network page Remote Tasks panel with near-realtime status updates and completion toasts.
2026-09-06 π v0.6.0 β Mesh operations makes the v0.5 mesh stack operationally usable: remote tasks are now exposed over the daemon gateway, the launcher, the web console, and the CLI, with capability-driven routing (active_tasks load), rhizome mesh route, mesh audit log, and new protocol unit tests. See the v0.6.0 release notes.
2026-09-05 π v0.5.0 β Mesh maturity adds NAT traversal (AutoNATv2, hole punching, circuit relay v2 with self-organizing relays), asynchronous remote agent tasks over /rhizome/agent-task/1.0.0, enforced mesh security (signed requests and capability manifests, per-peer ACLs, rate limits, audit trail), and hardened workspace sync with reconnect catch-up. See the v0.5.0 release notes.
2026-09-04 π v0.4.2βv0.4.8 shipped the live network status API and Network dashboard, trust/capability persistence, saved-peer management across CLI/daemon/launcher, and a catalog-driven provider protocol refactor with first-class local providers (Ollama, vLLM, LM Studio, LiteLLM). Release notes: v0.4.2βv0.4.8.
2026-09-01 π v0.4.0 β 32-bit and Android portability adds linux/386, windows/386, linux/armv7, and native Android arm64/arm/386/amd64 builds. P2P transports are hardened for Linux 3.4 / Android 4.4 kernels, and mDNS now fails gracefully when multicast is unavailable.
2026-05-28 π v0.2.9 Released! MCP server management in Web UI, configurable Sogou-backed web search, tool feedback animation in channels, pretty_print and disable_escape_html defaults, and numerous bug fixes across providers and channels.
2026-05-14 π v0.2.8 Released! MCP CLI commands (show, add, list, remove, test, edit), empty object instead of null for MCP tool parameters, and build fixes.
2026-05-07 π v0.2.7 Released! Configurable Sogou-backed web search, channel tool feedback animation, linter fixes.
2026-04-23 π v0.2.6 Released! Hooks with respond action and comprehensive documentation, isolation support, help banner fix.
2026-04-11 π v0.2.5 Released! Zoneinfo from TZ/ZONEINFO env, Matrix CommonMark rendering alignment, read_file by lines.
2026-03-31 π± Android Support! Rhizome now runs on Android! The Android APK is not currently distributed from this fork; build from source or check GitHub Releases for a future APK.
2026-03-25 π v0.2.4 Released! Agent architecture overhaul (SubTurn, Hooks, Steering, EventBus), WeChat/WeCom integration, security hardening (.security.yml, sensitive data filtering), new providers (AWS Bedrock, Azure, Xiaomi MiMo), and 35 bug fixes. Rhizome has reached 26K Stars!
2026-03-17 π v0.2.3 Released! System tray UI (Windows & Linux), sub-agent status query (spawn_status), experimental Gateway hot-reload, Cron security gating, and 2 security fixes. Rhizome has reached 25K Stars!
2026-03-09 π v0.2.1 β Biggest update yet! MCP protocol support, 4 new channels (Matrix/IRC/WeCom/Discord Proxy), 3 new providers (Kimi/Minimax/Avian), vision pipeline, JSONL memory store, model routing.
2026-02-28 π¦ v0.2.0 released with Docker Compose and Web UI Launcher support.
Earlier news...
2026-02-26 π Rhizome hits 20K Stars in just 17 days! Channel auto-orchestration and capability interfaces are live.
2026-02-16 π Rhizome breaks 12K Stars in one week! Community maintainer roles and Roadmap officially launched.
2026-02-13 π Rhizome breaks 5000 Stars in 4 days! Project roadmap and developer groups in progress.
2026-02-09 π Rhizome Released! Built in 1 day to explore ultra-lightweight AI Agents. Let's Go, Rhizome!
πͺΆ Single binary, no runtime dependencies: One statically-linked Go executable that runs on Linux, Windows, macOS, FreeBSD/NetBSD, and Android.*
π° Runs on what you own: Designed to live resident on hardware you already have β old Android phones, aging laptops, ARM/RISC-V boards, free-tier VMs β without getting in the way; see the Hardware Compatibility List.
β‘οΈ Lightning-fast boot: Starts in under a second on the verified low-cost boards.
π Truly portable: Single binary across RISC-V, ARM (including 32-bit ARMv7), MIPS, and x86 (including 32-bit i386) architectures. One binary, runs everywhere!
π€ AI-bootstrapped: Pure Go native implementation β 95% of core code was generated by an Agent and fine-tuned through human-in-the-loop review.
π MCP support: Native Model Context Protocol integration β connect any MCP server to extend Agent capabilities.
ποΈ Vision pipeline: Send images and files directly to the Agent β automatic base64 encoding for multimodal LLMs.
π§ Smart routing: Rule-based model routing β simple queries go to lightweight models, saving API costs.
*Measured on Windows with CGO_ENABLED=0, tags goolm,stdjson, and -ldflags "-s -w"; the stripped binary is ~74 MB.
| Mode | Use case | Total RAM | Free RAM | Storage |
|---|---|---|---|---|
| Base | One-shot rhizome agent, rhizome onboard |
256 MB | 128 MB | 128 MB |
| Full | rhizome daemon with P2P, syncer, and gateway |
512 MB | 256 MB | 128 MB |
Hardware Compatibility List β See all verified boards, from Raspberry Pi to Android phones. Your board not listed? Submit a PR!
Full-Stack Engineer Mode |
Logging & Planning |
Web Search & Learning |
|---|---|---|
| Develop Β· Deploy Β· Scale | Schedule Β· Automate Β· Remember | Discover Β· Insights Β· Trends |
Rhizome can be deployed on a wide range of Linux devices β including hardware you might otherwise retire!
- $15 Raspberry Pi Zero (or Zero 2 W), for a minimal home assistant
- $50~70 CanMV-K230, for RISC-V-based embedded use
- $100 NanoKVM-Pro, for automated server operations
- $100 MaixCAM2, for smart surveillance
See the Hardware Compatibility List for the full list of verified boards and the current two-tier requirements.
picoclaw_detect_person.mp4
π More Deployment Cases Await!
Visit the GitHub Releases page and download the binary for your platform.
Alternatively, download the binary for your platform from the GitHub Releases page.
Prerequisites:
- Go 1.26+
- Node.js 22+ and pnpm 10.33.0+ for Web UI / launcher builds
git clone https://github.com/stpinkie/rhizome.git
cd rhizome
make deps
# Install frontend dependencies
(cd web/frontend && pnpm install --frozen-lockfile)
# Build the core binary for the current platform
make build
# Build the Web UI Launcher (required for WebUI mode)
make build-launcher
# Build core binaries for all Makefile-managed platforms
make build-all
# Build for Raspberry Pi Zero 2 W
# 32-bit: make build-linux-arm
# 64-bit: make build-linux-arm64
make build-pi-zero
# Build and install
make installRaspberry Pi Zero 2 W: Use the binary that matches your OS: 32-bit Raspberry Pi OS -> make build-linux-arm; 64-bit -> make build-linux-arm64. Or run make build-pi-zero to build both.
The WebUI Launcher provides a browser-based interface for configuration and chat. This is the easiest way to get started β no command-line knowledge required.
Option 1: Double-click (Desktop)
After downloading from GitHub Releases, double-click rhizome-launcher (or rhizome-launcher.exe on Windows). Your browser will open automatically at http://localhost:18800.
Option 2: Command line
rhizome-launcher
# Open http://localhost:18800 in your browserTip
Remote access / Docker / VM: Add the -public flag to listen on all interfaces:
rhizome-launcher -publicGetting started:
Open the WebUI, then: 1) Configure a Provider (add your LLM API key) -> 2) Configure a Channel (e.g., Telegram) -> 3) Start the Gateway -> 4) Chat!
For detailed documentation, see the docs/ folder in this repo.
Docker (alternative)
# 1. Clone this repo
git clone https://github.com/stpinkie/rhizome.git
cd rhizome
# 2. First run β auto-generates docker/data/config.json then exits
# (only triggers when both config.json and workspace/ are missing)
docker compose -f docker/docker-compose.yml --profile launcher up
# The container prints "First-run setup complete." and stops.
# 3. Set your API keys
vim docker/data/config.json
# 4. Start
docker compose -f docker/docker-compose.yml --profile launcher up -d
# Open http://localhost:18800Docker / VM users: The Gateway listens on
127.0.0.1by default. SetRHIZOME_GATEWAY_HOST=0.0.0.0or use the-publicflag to make it accessible from the host.
# Check logs
docker compose -f docker/docker-compose.yml logs -f
# Stop
docker compose -f docker/docker-compose.yml --profile launcher down
# Update
docker compose -f docker/docker-compose.yml pull
docker compose -f docker/docker-compose.yml --profile launcher up -dmacOS β First Launch Security Warning
macOS may block rhizome-launcher on first launch because it is downloaded from the internet and not notarized through the Mac App Store.
Step 1: Double-click rhizome-launcher. You will see a security warning:
"rhizome-launcher" Not Opened β Apple could not verify "rhizome-launcher" is free of malware that may harm your Mac or compromise your privacy.
Step 2: Open System Settings β Privacy & Security β scroll down to the Security section β click Open Anyway β confirm by clicking Open Anyway in the dialog.
After this one-time step, rhizome-launcher will open normally on subsequent launches.
Give your decade-old phone a second life! Turn it into a smart AI Assistant with Rhizome.
Option 1: APK Install
Preview:
The Android APK is not currently published from this fork; build from source or check GitHub Releases for a future APK.
Option 2: Termux
For a full command-line setup checklist, see the Android Termux Guide.
Terminal Launcher (for resource-constrained environments)
- Install Termux (download from GitHub Releases, or search in F-Droid / Google Play)
- Run the following commands:
# Download the latest release
wget https://github.com/stpinkie/rhizome/releases/latest/download/rhizome_Linux_arm64.tar.gz
tar xzf rhizome_Linux_arm64.tar.gz
pkg install proot
termux-chroot ./rhizome onboard # chroot provides a standard Linux filesystem layoutThen follow the Terminal Launcher section below to complete configuration.
For minimal environments where only the rhizome core binary is available (no Launcher UI), you can configure everything via the command line and a JSON config file.
1. Initialize
rhizome onboardThis creates ~/.rhizome/config.json and the workspace directory.
2. Configure (~/.rhizome/config.json)
{
"agents": {
"defaults": {
"model_name": "gpt-5.4"
}
},
"model_list": [
{
"model_name": "gpt-5.4",
"model": "openai/gpt-5.4"
// api_key is now loaded from .security.yml
}
]
}See
config/config.example.jsonin the repo for a complete configuration template with all available options.Please note: config.example.json format is version 0, with sensitive codes in it, and will be auto migrated to version 1+, then, the config.json will only store insensitive data, the sensitive codes will be stored in .security.yml, if you need manually modify the codes, please see
docs/security/security_configuration.mdfor more details.
3. Chat
# One-shot question
rhizome agent -m "What is 2+2?"
# Interactive mode
rhizome agent
# Start gateway for chat app integration
rhizome gatewayRhizome supports 30+ LLM providers through the model_list configuration. Use the protocol/model format:
| Provider | Protocol | API Key | Notes |
|---|---|---|---|
| OpenAI | openai/ |
Required | GPT-5.4, GPT-4o, o3, etc. |
| Anthropic | anthropic-messages/ |
Required | Native Claude Messages API |
| Google Gemini | gemini/ |
Required | Gemini 3 Flash, 2.5 Pro, etc. |
| OpenRouter | openrouter/ |
Required | 200+ models, unified API |
| Zhipu (GLM) | zhipu/ |
Required | GLM-4.7, GLM-5, etc. |
| DeepSeek | deepseek/ |
Required | DeepSeek-V3, DeepSeek-R1 |
| Volcengine | volcengine/ |
Required | Doubao, Ark models |
| Qwen | qwen/ |
Required | Qwen3, Qwen-Max, etc. |
| Groq | groq/ |
Required | Fast inference (Llama, Mixtral) |
| Moonshot (Kimi) | moonshot/ |
Required | Kimi models |
| Minimax | minimax/ |
Required | MiniMax models |
| Mistral | mistral/ |
Required | Mistral Large, Codestral |
| NVIDIA NIM | nvidia/ |
Required | NVIDIA hosted models |
| Cerebras | cerebras/ |
Required | Fast inference |
| NEAR AI Cloud | nearai/ |
Required | TEE inference, OpenAI-compatible |
| Novita AI | novita/ |
Required | Various open models |
| Xiaomi MiMo | mimo/ |
Required | MiMo models |
| Ollama | ollama/ |
Not needed | Local models, self-hosted |
| vLLM | vllm/ |
Not needed | Local deployment, OpenAI-compatible |
| LM Studio | lmstudio/ |
Not needed | Local GUI server, OpenAI-compatible |
| LiteLLM | litellm/ |
Varies | |
| Azure OpenAI | azure/ |
API key or Entra ID** | Enterprise Azure deployment |
| GitHub Copilot | github-copilot/ |
OAuth | Device code login |
| Antigravity | antigravity/ |
OAuth | Google Cloud AI |
| AWS Bedrock* | bedrock/ |
AWS credentials | Claude, Llama, Mistral on AWS |
* AWS Bedrock requires build tag:
go build -tags bedrock. Setapi_baseto a region name (e.g.,us-east-1) for automatic endpoint resolution across all AWS partitions (aws, aws-cn, aws-us-gov). When using a full endpoint URL instead, you must also configureAWS_REGIONvia environment variable or AWS config/profile.** Azure OpenAI uses
api_keywhen set. Ifapi_keyis omitted, the provider falls back to Microsoft Entra ID viaDefaultAzureCredential(env vars, workload identity, managed identity, Azure CLI, etc.). The Entra ID path requires build tag:go build -tags azidentity.
Local deployment (Ollama, vLLM, etc.)
Ollama:
{
"model_list": [
{
"model_name": "local-llama",
"model": "ollama/llama3.1:8b",
"api_base": "http://localhost:11434/v1"
}
]
}vLLM:
{
"model_list": [
{
"model_name": "local-vllm",
"model": "vllm/your-model",
"api_base": "http://localhost:8000/v1"
}
]
}LM Studio:
{
"model_list": [
{
"model_name": "local-lmstudio",
"model": "lmstudio/your-model",
"api_base": "http://localhost:1234/v1"
}
]
}For full provider configuration details, see Providers & Models.
Talk to your Rhizome through 19+ messaging platforms:
| Channel | Setup | Protocol | Docs |
|---|---|---|---|
| Telegram | Easy (bot token) | Long polling | Guide |
| Discord | Easy (bot token + intents) | WebSocket | Guide |
| Easy (QR scan or bridge URL) | Native / Bridge | Guide | |
| Weixin | Easy (Native QR scan) | iLink API | Guide |
| Easy (AppID + AppSecret) | WebSocket | Guide | |
| Slack | Easy (bot + app token) | Socket Mode | Guide |
| Matrix | Medium (homeserver + token) | Sync API | Guide |
| Delta Chat | Easy (account script or email/password) | JSON-RPC (email/E2EE) | Guide |
| Feishu / Lark | Medium (App ID + Secret) | WebSocket/SDK | Guide |
| LINE | Medium (credentials + webhook) | Webhook | Guide |
| WeCom | Easy (QR login or manual) | WebSocket | Guide |
| VK | Easy (group token) | Long Poll | Guide |
| IRC | Medium (server + nick) | IRC protocol | Guide |
| OneBot | Medium (WebSocket URL) | OneBot v11 | Guide |
| MQTT | Easy (broker + agent_id) | < 1798 td>MQTT pub/subGuide | |
| MaixCam | Easy (enable) | TCP socket | Guide |
| Pico | Easy (enable) | Native protocol | Built-in |
| Pico Client | Easy (WebSocket URL) | WebSocket | Built-in |
All webhook-based channels share a single Gateway HTTP server (
gateway.host:gateway.port, default127.0.0.1:18790). Feishu uses WebSocket/SDK mode and does not use the shared HTTP server.
Log verbosity is controlled by
gateway.log_level(default:warn). Supported values:debug,info,warn,error,fatal. Can also be set viaRHIZOME_LOG_LEVEL. See Configuration for details.
For detailed channel setup instructions, see Chat Apps Configuration.
Rhizome can search the web to provide up-to-date information. Configure in tools.web:
| Search Engine | API Key | Free Tier | Link |
|---|---|---|---|
| DuckDuckGo | Not needed | Unlimited | Built-in fallback |
| Gemini Google Search | Required | Varies | Gemini with Google Search grounding |
| Baidu Search | Required | 1500/month (daily allocation) | AI-powered, China-optimized |
| Tavily | Required | 1000 queries/month | Optimized for AI Agents |
| Brave Search | Required | 2000 queries/month | Fast and private |
| Kagi Search | Required | Paid/limited by API setup | Premium search results |
| Perplexity | Required | Paid | AI-powered search |
| SearXNG | Not needed | Self-hosted | Free metasearch engine |
| GLM Search | Required | Varies | Zhipu web search |
Rhizome includes built-in tools for file operations, code execution, scheduling, and more. See Tools Configuration for details.
Skills are modular capabilities that extend your Agent. They are loaded from SKILL.md files in your workspace.
Install skills from ClawHub:
rhizome skills search "web scraping"
rhizome skills install <skill-name>Configure skill registries:
Add to your config.json:
{
"tools": {
"skills": {
"registries": {
"clawhub": {
"auth_token": "your-clawhub-token"
},
"github": {
"base_url": "https://github.com",
"auth_token": "your-github-token",
"proxy": ""
}
}
}
}
}tools.skills.github.* is deprecated. Use tools.skills.registries.github.* instead.
For more details, see Tools Configuration - Skills.
Rhizome natively supports MCP β connect any MCP server to extend your Agent's capabilities with external tools and data sources.
{
"tools": {
"mcp": {
"enabled": true,
"servers": {
"filesystem": {
"enabled": true,
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem", "/tmp"]
}
}
}
}
}You can manage common MCP setups directly from the CLI instead of editing JSON by hand:
rhizome mcp add filesystem -- npx -y @modelcontextprotocol/server-filesystem /tmp
rhizome mcp list
rhizome mcp test filesystemrhizome mcp is a configuration manager: it updates config.json under tools.mcp.servers, but it does not keep the server process running itself.
Use rhizome mcp edit when you need advanced fields that are not covered by rhizome mcp add.
For example, rhizome mcp add supports --deferred and --env-file, while rhizome mcp edit is still useful for direct JSON editing and uncommon MCP settings.
For full MCP configuration (stdio, SSE, HTTP transports, Tool Discovery), see Tools Configuration - MCP. For CLI usage and examples, see MCP Server CLI.
Connect Rhizome to the Agent Social Network simply by sending a single message via the CLI or any integrated Chat App.
Read https://clawdchat.ai/skill.md and follow the instructions to join ClawdChat.ai
| Command | Description |
|---|---|
rhizome onboard |
Initialize config & workspace |
rhizome auth weixin |
Connect WeChat account via QR |
rhizome agent -m "..." |
Chat with the agent |
rhizome agent |
Interactive chat mode |
rhizome gateway |
Start the gateway |
rhizome status |
Show status |
rhizome version |
Show version info |
rhizome model |
View or switch the default model |
rhizome mcp list |
List configured MCP servers |
rhizome mcp add ... |
Add or update an MCP server entry |
rhizome mcp test |
Probe a configured MCP server |
rhizome mcp edit |
Open config for advanced MCP editing |
rhizome mcp remove |
Remove an MCP server entry |
rhizome cron list |
List all scheduled jobs |
rhizome cron add ... |
Add a scheduled job |
rhizome cron disable |
Disable a scheduled job |
rhizome cron remove |
Remove a scheduled job |
rhizome skills list |
List installed skills |
rhizome skills install |
Install a skill |
rhizome migrate |
Migrate data from older versions |
rhizome auth login |
Authenticate with providers |
Rhizome supports scheduled reminders and recurring tasks through the cron tool:
- One-time reminders: "Remind me in 10 minutes" -> triggers once after 10min
- Recurring tasks: "Remind me every 2 hours" -> triggers every 2 hours
- Cron expressions: "Remind me at 9am daily" -> uses cron expression
See docs/reference/cron.md for current schedule types, execution modes, command-job gates, and persistence details.
For detailed guides beyond this README:
| Topic | Description |
|---|---|
| Docker & Quick Start | Docker Compose setup, Launcher/Agent modes |
| Chat Apps | All 18+ channel setup guides |
| Configuration | Environment variables, workspace layout, security sandbox |
| MCP Server CLI | Add, list, test, edit, and remove MCP server entries from the CLI |
| Scheduled Tasks and Cron Jobs | Cron schedule types, deliver modes, command gates, job storage |
| Providers & Models | 30+ LLM providers, model routing, model_list configuration |
| Spawn & Async Tasks | Quick tasks, long tasks with spawn, async sub-agent orchestration |
| Hooks | Event-driven hook system: observers, interceptors, approval hooks |
| Steering | Inject messages into a running agent loop between tool calls |
| SubTurn | Subagent coordination, concurrency control, lifecycle |
| Troubleshooting | Common issues and solutions |
| Tools Configuration | Per-tool enable/disable, exec policies, MCP, Skills |
| Hardware Compatibility | Tested boards, minimum requirements |
PRs welcome! The codebase is intentionally small and readable.
See our Community Roadmap and CONTRIBUTING.md for guidelines.
Developer group building, join after your first merged PR!
User Groups:
Discord: https://discord.gg/V4sAZ9XWpN