A Chinese-language tutorial series for installing, configuring and building your own OpenClaw AI assistant, from beginner setup to custom Skills development.
nikships/muse-cli
Command-line client for a personal muse.ai AI agent: chat, side chats, feed, goals, ideas and sessions from the terminal.
- Stars
- 51
- Forks
- 20
- Open issues
- 1
- Last commit
- 14 hours ago
What is it?
muse-cli lets you talk to your personal muse.ai agent without opening the browser. It covers chatting, side chats, the feed, goals, ideas and session management through named subcommands, and a raw command reaches any of the 258 gateway methods. Each command prints JSON, so output can be piped into jq, cron jobs or other scripts.
The tool connects the way the muse.ai web app does: it authenticates over HTTPS, then opens an encrypted Noise-XX WebSocket to the user's personal VM. Login reuses cookies from Google Chrome, which are exported once with the agent-browser helper and stored locally with restricted file permissions. Access tokens are fetched fresh for each command.
The project is written in Python, installs with uv, pipx or pip, and includes an agent skill so coding agents can run read-only muse-cli commands on the user's behalf. Because the internal APIs are unversioned, the project notes that calls may break when muse.ai changes.
Key features
- 01Chat with the agent and read replies as JSON from the shell.
- 02Manage side chats: list, rename, pin, archive and delete sessions.
- 03Read and react to the feed, and view goals and ideas.
- 04Run a raw escape hatch for any of the 258 gateway methods.
- 05Authenticate by reusing Chrome cookies, saved locally with mode 600.
- 06Ships an agent skill so coding agents can run commands for you.
Use cases
- Piping agent replies into shell scripts or jq pipelines.
- Scheduling recurring agent queries with cron.
- Letting a coding agent check your status and goals from the terminal.
- Using a terminal-based or screen-reader-friendly chat interface.
Quick start
- 1
Install the CLI with uv
uv tool install muse-cli - 2
Install the cookie reader from npm (needs Node.js)
npm i -g agent-browser - 3
Turn on remote debugging in Chrome, log in on muse.ai, then save the login
- 4
Check the connection
muse-cli status
Things to know
- Relies on borrowed browser cookies from Google Chrome and unversioned internal muse.ai APIs that can change.
- Requires a muse.ai account and a personal VM; cookies expire and need repeating the login steps.
- Depends on the Chrome remote debugging setup and agent-browser for the easiest login path.
Topics
FAQ
- What is muse-cli?
- Command-line client for a personal muse.ai AI agent: chat, side chats, feed, goals, ideas and sessions from the terminal. muse-cli lets you talk to your personal muse.ai agent without opening the browser. It covers chatting, side chats, the feed, goals, ideas and session management through named subcommands, and a raw command reaches any of the 258 gateway methods. Each command prints JSON, so output can be piped into jq, cron jobs or other scripts.
- Is muse-cli free to use? What license does it have?
- muse-cli is open source under the MIT license. Check the license terms before using it in commercial products.
- Is muse-cli still maintained?
- Actively maintained. The last commit to muse-cli was on Oct 22, 2026.
- What can I use muse-cli for?
- Piping agent replies into shell scripts or jq pipelines. Scheduling recurring agent queries with cron. Letting a coding agent check your status and goals from the terminal. Using a terminal-based or screen-reader-friendly chat interface.
- How do I install muse-cli?
- Install the CLI with uv: uv tool install muse-cli Install the cookie reader from npm (needs Node.js): npm i -g agent-browser Turn on remote debugging in Chrome, log in on muse.ai, then save the login Check the connection: muse-cli status
- Where can I get muse-cli?
- The source code and installation instructions are on GitHub: https://github.com/nikships/muse-cli
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A self-hosted personal AI assistant that runs on your own devices and answers you through messaging apps like WhatsApp, Telegram, Slack and Discord.
Open-source multi-agent AI assistant that runs tasks across web, desktop, files and chat apps like Telegram, Feishu and WeCom.
A self-improving AI agent that creates and refines skills from experience, runs on many model providers, and works through CLI and messaging apps.
An open-source personal AI assistant runtime in Go that takes requests over email, SMS, XMPP, and HTTP and acts through tools like mail, calendar, and files.
Self-hosted personal AI agent framework in Python with a browser WebUI, terminal client, chat app channels, tools, memory and MCP support.