Introduction

What is CoPaw?

CoPaw is a personal AI workstation. It is meant for people who want an assistant they can keep inside their own environment, connect to real communication channels, and operate over time instead of restarting from zero every day.

The product sits between two familiar categories:

• Frameworks give you raw power but require you to build the product layer yourself.
• Closed assistants give you polish but very little ownership.

CoPaw tries to combine ownership with product surface. You get a runtime, a browser Console, install paths, Skills, channel setup, and docs that explain how to keep the system useful.

How do people use it?

Most people touch CoPaw in two loops:

Direct interaction

You talk to it in a channel or through the browser Console. That can mean:

• asking a question;
• summarizing a document;
• running a Skill-based workflow;
• switching providers or models;
• checking context or token usage.

Ongoing operation

The more interesting use case is not the one-off answer. It is the repeat loop:

• scheduled digests;
• reminders pushed to a real chat app;
• recurring research runs;
• personal workflow check-ins;
• long-lived memory and workspace context.

What makes it different?

Three things matter most:

1. The channel layer is first-class

CoPaw is not designed around a single proprietary chat window. It connects to messaging surfaces such as DingTalk, Feishu, QQ, Discord, iMessage, Telegram, Matrix, Mattermost, and other adapters.

2. Behavior is extendable through Skills

Built-in and workspace-defined Skills let you broaden what the assistant can do without locking yourself into one vendor-defined interface.

3. The operating surface is part of the product

The browser Console, CLI, docs, and deployment paths are part of the workstation story. They are not side notes for contributors only.

Suggested reading order

If you are new, the shortest useful path is:

1. Read Quick Start
2. Open Console
3. Connect Channels
4. Review Skills
5. Understand Heartbeat

A note on deployment

If you want to self-serve, the docs in this site are enough to get started.

If your goal is a traffic-ready or production-assisted deployment, this site routes deployment-oriented CTA traffic to EasyClaw so that operational help lands in the right place.