Claw Spark — AI skill for Claude Code
One-click AI agent setup for NVIDIA DGX Spark, Jetson, and RTX hardware.
How to install Claw Spark
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open rohitg00/claw-spark and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claw Spark does
One-click AI agent setup for NVIDIA DGX Spark, Jetson, and RTX hardware. OpenClaw + Ollama, fully local.
Alternatives in AI
- Osaurus — Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cry 5.1k ★
- Manifest — Real-time cost observability for OpenClaw agents — track tokens, costs, messages, and model usage 4.1k ★
- EchoBird — One-click install + model switch:Claude Code,Codex CLI (OpenAI), Grok Build (xAI), DeepSeek Harness, Kimi Code 3.1k ★
README
___ _ ___ __ __ ___ ___ ___ ___ _ __
/ __|| | / \ \ \ / / / __|| _ \ / \ | _ \| |/ /
| (__ | |__ | (_) | \ \/\/ / \__ \| _/| (_) || /| ' <
\___||____| \___/ \_/\_/ |___/|_| \___/ |_|_\|_|\_\
clawspark
**One command. Private AI agent. Your hardware.**
curl -fsSL https://clawspark.dev/install.sh | bash
That's it. Go grab a coffee. Come back to a fully working, fully private AI agent that can code, research, browse the web, analyze images, and manage your tasks -- all running on your own hardware.
What is this?
[OpenClaw](https://github.com/openclaw/openclaw) is the most popular open-source AI agent on the planet (247K+ stars). **clawspark** gets it running on your NVIDIA hardware in one command. Fully local. Fully private. Your data never leaves your machine. No cloud APIs, no subscriptions, no telemetry.
What happens when you run it
- Detects your hardware (DGX Spark, Jetson, RTX GPUs, Mac)
- Picks the best model using llmfit for hardware-aware selection
- Installs everything (Ollama, OpenClaw, skills, dependencies)
- Configures multi-model (chat model + optional vision model for image analysis)
- Enables voice (local Whisper transcription, zero cloud)
- Sets up browser automation (headless Chromium for web tasks)
- Sets up your dashboard (built-in chat UI + ClawMetry metrics)
- Creates systemd services (auto-starts on boot, survives reboots)
- Hardens security (firewall rules, auth tokens, localhost binding, optional Docker sandbox)
Total time: about 5 minutes on DGX Spark with a decent connection.
Supported Hardware
**Tested and verified:**
| Hardware | Memory | Default Model | Tokens/sec |
|---|---|---|---|
| DGX Spark | 128 GB unified | Qwen 3.5 35B-A3B | ~59 (measured) |
DGX Spark has a curated model list tested on real hardware.
**Should work (community testing welcome
Related Skills
HomeAILab
Home AI inference lab running Qwen3.8 across RTX 5090, RTX 3090, and DGX Spark with vLLM, llama.cpp, custom Go
Agent Observability Stack
Self-hostable observability for LLM agents + a Linux host (Intel iGPU/NPU + NVIDIA RTX 3090 eGPU aware): Prome
Local LLM Setup
Runbook and PowerShell automation to host two isolated local LLMs (KYC analyst and KB-chatbot reasoning) on on
Long Run Agent
🔧 LRA - AI Agent长任务超稳定执行框架 ⚡ 轻量 pip一键安装|突破上下文限制 📌 7种状态管理 Long-Running AI Agent Framework • One-click pip ins
Local LLM Agentic Coding
Install a free local LLM (Ollama) and connect Claude Code or OpenCode to it, with an honest hardware reality-c
Claw0
0 - 1 learn OpenClaw: sections to build an claw-AI agent from scratch
Related Agents
Lemonade Specialist
Lemonade Server and SDK specialist for local LLM on AMD hardware. Use PROACTIVELY for Lemonade setup, model ma
Code Porter
Feature porting specialist for the hl_claw_bot two-repo setup. Extracts and upgrades features from the game-ba
Hardware Engineer
Firmware and embedded work for projects with a hardware component — PlatformIO builds, device-side logic, and