by Nexting-ai
A wearable terminal to dispatch tasks to your own AI agents — Claude Code, OpenClaw, Codex. Tap, speak, dispatch. The first hardware built for OpenClaw. Open source (MIT).
# Add to your Claude Code skills
git clone https://github.com/Nexting-ai/nextingLast scanned: 7/9/2026
{
"issues": [
{
"file": "README.md",
"line": 87,
"type": "remote-install",
"message": "Install command (remote install script piped to a shell — review the source before running): \"curl -fsSL https://nexting.ai/install-cc | bash\"",
"severity": "low"
}
],
"status": "PASSED",
"scannedAt": "2026-07-09T07:46:52.852Z",
"npmAuditRan": true,
"pipAuditRan": true,
"promptInjectionRan": true
}nexting is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by Nexting-ai. A wearable terminal to dispatch tasks to your own AI agents — Claude Code, OpenClaw, Codex. Tap, speak, dispatch. The first hardware built for OpenClaw. Open source (MIT). It has 108 GitHub stars.
Yes. nexting passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/Nexting-ai/nexting" and add it to your Claude Code skills directory (see the Installation section above).
nexting is primarily written in C. It is open-source under Nexting-ai on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh nexting against similar tools.
No comments yet. Be the first to share your thoughts!
You already have powerful agents — Claude Code, Codex, OpenClaw. They're locked to your desk. Nexting is the channel between you and them: a terminal you wear, so you can reach your own agent without pulling out your phone or opening an app. Tap it, say one sentence, and the task is dispatched to your own agent. It runs in the background and the result comes back when it's done.
Dispatch, not chat — press, speak, move on.
Nexting comes in two form factors, one capability. The PIN ($129, shipping now) pins to your collar — the one you can get today. The Ring — our flagship, in private beta — makes "raise your hand and dispatch an agent" as invisible as a ring on your finger. Same capabilities, two shapes. Your whole agent team, always on you.
You speak → Nexting PIN (BLE) → iPhone App (STT) → Cloud → your own agent
(Claude Code / Codex / OpenClaw)
↓
Result pushed back ← iPhone App ← ── ── agent runs it in the background
3 seconds to dispatch a task — vs. ~30s pulling out your phone, unlocking, and typing.
| Dispatch, not chat | Fire-and-forget: say it once, your agent runs it in the background. Not Q&A. |
| Your own agent | Drives the agents you already run — Claude Code, Codex, OpenClaw. Not a locked-in model. |
| Private by design | Claude Code & Codex support optional end-to-end encryption — turn it on and Nexting relays ciphertext, not your session. |
| Delivers while you're offline | Phone locked, on the move? The agent still finishes the job and pushes the result back. |
| Remote-control your agent | Attach to a Claude Code or Codex session on your Mac or Windows PC and drive it from your pocket — live. |
| Deep iPhone integration | Calendar, Reminders, Contacts, Health, HomeKit — all by voice. |
Nexting isn't another voice assistant — it's the pocket control surface for your agent team.
Nexting is a terminal, not a model. Connect the agent you already run:
| Mode | What It Is | Cost |
|---|---|---|
| Claude Code | Attach to a Claude Code session on your Mac or Windows PC and drive it from your PIN — live. | Free |
| Codex | Attach to an OpenAI Codex CLI session on your Mac or Windows PC and drive it from your PIN — live. | Free |
| MyOpenClaw | Run your own OpenClaw instance. We handle the relay. | Free |
| MyHermes | Any OpenAI-compatible local AI — Hermes Agent, Ollama, vLLM, LM Studio. | Free |
| Nexting Pro | Managed agent in the cloud. Latest Claude, GPT, and Gemini models, zero setup. | $29/mo or $279/yr |
Buy the hardware once. Dispatch to whichever agent is yours. Expand your mode below for setup — each takes about a minute.
One command on the computer where Claude Code runs. It opens your browser to sign in to Nexting — that login binds the machine to your account — and installs everything, including a background daemon that keeps your phone connected.
macOS
curl -fsSL https://nexting.ai/install-cc | bash
Windows (PowerShell)
irm https://nexting.ai/install-cc.ps1 | iex
Then just run claude as usual (on Windows, fully quit and reopen your terminal first) — your sessions appear in the Nexting app, live. Uninstall anytime: nexting-cc-bridge uninstall.
One command to sign in and install:
macOS
curl -fsSL https://nexting.ai/install-codex | bash
Windows (PowerShell)
irm https://nexting.ai/install-codex.ps1 | iex
Then just run codex as usual — your sessions appear in the Nexting app, live. Uninstall anytime: nexting-cc-bridge codex-uninstall.
Nexting is the first hardware product built for OpenClaw. Install the plugin and link your account:
openclaw plugins install pinclaw --dangerously-force-unsafe-install
openclaw pinclaw login
The npm package and CLI commands use the internal codename
pinclaw— that's by design, not a leftover. Full plugin docs:plugin/README.md.
Point Nexting at your local endpoint with the nexting-hermes-bridge CLI:
npm install -g nexting-hermes-bridge
nexting-hermes-bridge login
nexting-hermes-bridge start --endpoint http://localhost:8642 --model hermes-agent
Subscribe in the Nexting app and it's ready to dispatch.
Full walkthrough for every mode — pairing, remote control, starting sessions by voice: nexting.ai/doc.
Purpose-built for voice-first interaction. No screen — by design.
| Spec | Detail |
|---|---|
| MCU | Seeed XIAO nRF52840 Sense (ARM Cortex-M4 @ 64MHz) |
| Microphone | PDM MEMS (built into XIAO Sense) |
| Audio | Opus codec over BLE 5.0, I2S speaker (MAX98357A) |
| Feedback | RGB LED + speaker (no screen) |
| Battery | 3.7V LiPo, USB-C charging (onboard BQ25101) |
| Firmware | Zephyr RTOS v2.2.0 (source + UF2) |
| Interaction | Single-button push-to-talk |
The hardware-opensource/ directory is a complete Co-Builder Edition: firmware source, 3D-printable enclosure files, and a schematic PDF — everything you need to build a Nexting PIN yourself. Flash the UF2 binary via drag-and-drop — no programmer needed.
Your phone isn't just a pipe — it's the bridge between you and your AI. The Nexting app registers native iOS capabilities as skills your agent can call:
Say "schedule a meeting tomorrow at 3pm" — the agent calls your iPhone's calendar directly.
All data stays on your iPhone. You control every permission.
nexting/
├── hardware-opensource/ # DIY Co-Builder Kit
│ ├── firmware/ # Zephyr firmware (source + UF2 binary)
│ ├── enclosure/ # 3D printing files (STL + OpenSCAD)
│ └── docs/ # Schematic PDF
├── plugin/ # OpenClaw channel plugin (npm package)
├── agent-bus/ # Multi-agent team bus (wake local agents from one Lark group)
└── public/ # Product assets
The codename
pinclawpersists in package names, file paths, and identifiers — it's the permanent internal handle. Nexting is the brand. Both coexisting is intentional.
| Website | nexting.ai |
| iOS App | App Store |
| Docs | nexting.ai/doc |
| Get a PIN | nexting.ai/reserve |
Issues and pull requests welcome.
MIT