aNother Claw
in your pocket

NClaw is a container-first AI assistant powered by Claude Code. A single Go binary. ~10 MB idle. Each Telegram topic becomes a separate project with its own session, files, and context. docker run and you're done.

Get Started View Source

nclaw

$ docker run ghcr.io/nickalie/nclaw

[10:00:01] scheduler: loaded 3 active tasks

[10:00:01] scheduler: started

[10:00:01] webhook: server listening on :3000

[10:00:01] nclaw bot started (version=0.5.2 commit=a3b1c9f)

[10:00:05] handler: received message chat=8241953

[10:00:12] pipeline: sending reply len=247

About

What is nclaw?

A lightweight, container-first AI assistant powered by Claude Code, accessible through Telegram. Each chat thread maintains its own session, and each Telegram topic becomes a separate project with isolated context and files. Send text, photos, documents, audio, or video. Claude runs inside a Docker container with full tool access, including git, gh CLI, Chromium, Go, Node.js, and Python. Results come back as formatted Telegram messages or file attachments.

Capabilities

Everything You Need

A full-featured AI assistant accessible from anywhere

Session Persistence

Each chat thread maintains its own Claude session. Pick up where you left off, even across restarts. Context is preserved automatically.

File Attachments

Send photos, documents, audio, video to Claude. Files are cached and referenced in prompts automatically.

Scheduled Tasks

Create recurring or one-time jobs using natural language. Tasks persist across restarts and run autonomously.

Rich Runtime

Docker image includes git, gh CLI, Chromium, Go, Node.js, and Python/uv. The assistant can install additional packages on the fly.

File Delivery

Claude can send files back to you — generated reports, exports, code artifacts, images, anything.

Telegram Topics

Each topic in a group chat is a separate project with isolated context, history, and files. One chat becomes a multi-project workspace.

Webhooks

Register HTTP endpoints that forward incoming requests to Claude. Connect GitHub events, package tracking, smart home alerts, and more.

HTML Formatting

Responses render using Telegram's HTML formatting with proper code blocks, inline markup, and a plain-text fallback for compatibility.

Workflow

How It Works

Simple architecture, powerful results

Message

Send a message, photo, document, or voice note to the nclaw bot in Telegram.

Invoke

nclaw launches Claude Code CLI inside a Docker container with your prompt.

Execute

Claude runs with full tool access — files, commands, browser, scripts.

Respond

Results are formatted and sent back to your Telegram chat with rich formatting.

Extensions

Built-in Skills

Six skills out of the box. The assistant can also create its own skills on the fly when a task requires specialized or repeatable behavior.

schedule Create and manage recurring or one-time scheduled tasks using natural language

send-file Deliver generated files — reports, exports, code artifacts — via Telegram

find-skills Discover and install additional skills to extend capabilities

agent-browser Browse the web using a headless Chromium instance inside the container

webhook Register HTTP endpoints that forward incoming requests to Claude

skill-creator Guide for creating new custom skills

Setup

Get Started

Clone, configure, deploy — have your personal Claude assistant in minutes

Docker

docker run -d --name nclaw \
  -e NCLAW_TELEGRAM_BOT_TOKEN=your-token \
  -e NCLAW_TELEGRAM_WHITELIST_CHAT_IDS=your-chat-id \
  -e NCLAW_DATA_DIR=/app/data \
  -v ./data:/app/data \
  -v ~/.claude/.credentials.json:/root/.claude/.credentials.json:ro \
  ghcr.io/nickalie/nclaw:latest

Kubernetes (Helm)

helm install nclaw \
  oci://ghcr.io/nickalie/charts/nclaw \
  --set env.telegramBotToken=your-token \
  --set env.whitelistChatIds=your-chat-id \
  --set claudeCredentialsSecret=my-secret

Full configuration guide →

Documentation

# everything you need to know

Read the docs →

aNother Clawin your pocket