DEDUPED Hot AI repos. Real use cases. What people build. Get the Monday 5

How to deploy Qwen Paw in Docker with data, secrets and backups kept with Qwen Paw

Run the official Docker image with separate volumes for data, secrets and backups.

Short answer: Run docker pull agentscope/qwenpaw:latest; then start it with the README's docker run command, mapping 127.0.0.1:8088 and the qwenpaw-data, qwenpaw-secrets and qwenpaw-backups volumes; open http://127.0.0.1:8088/ for the console. All steps.

Who it is forA home server owner who wants the assistant always on and easy to upgrade.

The problemRunning an assistant in a container loses memory and keys on every upgrade unless the volumes are set up right.

The outcomeQwen Paw's console on 127.0.0.1:8088 with config, memory and skills kept across restarts.

What you need first

  • Python 3.11 or newer for the pip install, or Docker
  • A model: QwenPaw Local with QwenPaw-Flash models, Ollama on the host, or a provider API key
  • Accounts for any chat channels you want to connect (Discord, Telegram and others)

Steps

  1. Run docker pull agentscope/qwenpaw:latest
  2. Start it with the README's docker run command, mapping 127.0.0.1:8088 and the qwenpaw-data, qwenpaw-secrets and qwenpaw-backups volumes
  3. Open http://127.0.0.1:8088/ for the console
  4. If your models run in Ollama on the host, use the README's host binding note because localhost inside the container is the container itself

Prompt for your agent

Deploy agentscope-ai/QwenPaw with Docker on <server> using the README's run command and the three named volumes, connect it to Ollama on the host, and confirm memory survives a container restart.

Caveat.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for agentscope-ai/QwenPaw and "Deploy Qwen Paw in Docker with data, secrets and backups kept"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer