AI CLI Agent Swarm: Run Multiple AI Coding Agents in Parallel
Run ten coding agents in parallel with CodeAgentSwarm, including Claude Code, Codex, Muse Code, Pi and Devin CLI. One workspace for every session.
Read guide
Guides
Guides
Learn how to get the most out of CodeAgentSwarm with our practical guides, grouped by the CLI you use.
Cross-CLI & Agent Swarm
Git Worktrees for AI Coding Agents: Give Each Agent Its Own Checkout
Running several AI agents on one repo means they fight over the same files. Git worktrees give each agent its own checkout on its own branch. Here is how.
Read guide
Git Worktree vs Branch for Parallel AI Agents: Which to Use
Switching branches, cloning the repo, or git worktrees? Here is why worktrees win for running parallel AI agents, and when a plain branch is still enough.
Read guide
The Best Tools to Run Multiple AI Coding Agents in Parallel (2026)
CodeAgentSwarm, T3 Code, Superset, Paseo, Conductor, Claude Squad and more compared: stars, last commit, platforms. Every fact verified on August 25, 2026.
Read guide
T3 Code vs CodeAgentSwarm: An Honest Comparison (2026)
T3 Code is an open source control plane that ends in a one-click PR. CodeAgentSwarm is a supervision workspace for ten agent CLIs. Honest 2026 comparison.
Read guide
Superset vs CodeAgentSwarm: An Honest Comparison (2026)
Superset is a source available code editor for running 10+ parallel agents. CodeAgentSwarm is a supervision workspace for ten agent CLIs. Honest 2026 comparison.
Read guide
Paseo vs CodeAgentSwarm: An Honest Comparison (2026)
Paseo is a self-hosted, open source agent orchestrator you can drive from your phone. CodeAgentSwarm is a macOS, Windows and Linux desktop workspace for ten agent CLIs.
Read guide
Conductor vs CodeAgentSwarm: An Honest Comparison (2026)
Compare Conductor local and cloud workspaces, four agent CLIs and current pricing with CodeAgentSwarm on macOS, Windows and Linux.
Read guide
Claude Squad vs CodeAgentSwarm: An Honest Comparison (2026)
Claude Squad manages parallel AI agents inside your terminal. CodeAgentSwarm is a desktop app for the same job. Honest comparison, verified August 2026.
Read guide
Vibe Kanban vs CodeAgentSwarm: An Honest Comparison (2026)
Vibe Kanban turns agent work into a kanban of issues. CodeAgentSwarm lets you watch and drive the agent terminals. Honest comparison with facts verified August 2026.
Read guide
Nimbalyst vs CodeAgentSwarm: An Honest Comparison (2026)
Nimbalyst, formerly Crystal, is a visual editor for what Claude Code and Codex produce. CodeAgentSwarm supervises ten agent CLIs. Honest comparison, verified in 2026.
Read guide
Claude Code
Claude Code Weekly Limit: A Daily Budget That Makes Your Quota Last All Week
Stop burning your Claude Code weekly limit in two days. Set a daily budget per provider, spread what is left until the reset, and pause new work at the limit.
Read guide
AI Coding Agent Coordinator: One Agent That Plans and Runs the Others
A coordinator agent splits your goal, opens worker sessions with the right agent and model for each job, and reports back in one Chat. Project and global modes.
Read guide
Auto Kanban for AI Coding Agents: Tasks That Start on Their Own
Drop tasks into the Auto lane and each one starts its own session with the agent, model, reasoning and permissions you chose. Finished work lands in In Testing.
Read guide
How to Run Multiple Claude Code Terminals in Parallel with CodeAgentSwarm
Run multiple Claude Code terminals in parallel with isolated sessions, live diffs and notifications. Includes setup, conflict prevention and practical workflows.
Read guide
Search and Resume Any Claude Code Conversation with CodeAgentSwarm History
Search Claude Code history across every project, preview conversations, and resume a session in one click with CodeAgentSwarm's local conversation history.
Read guide
Claude Code Notifications: Know When AI Finishes Without Watching the Terminal
Desktop alerts the moment Claude Code finishes or needs your input. Stop staring at the terminal: set it up in minutes and get your focus back.
Read guide
View Claude Code Changes in Real Time: Live Diff, Git Diff & AI Tracking
See exactly what Claude Code is changing in your code in real time. Three methods: session live diff per terminal, project-level Git diff viewer, and dynamic titles showing what each AI agent is doing.
Read guide
Claude Code History: Where It Is Stored and How to Find, Back Up & Resume It
Find Claude Code session files, search JSONL history, resume previous conversations, and back up or restore chats before automatic cleanup removes them.
Read guide
Can You Run Multiple Claude Code Sessions at Once? Yes, Here's How
Run multiple Claude Code sessions in parallel with terminal tabs, tmux, or CodeAgentSwarm. Compare isolation, visibility, conflicts, and the safest setup.
Read guide
How to Use Multiple Claude Code Accounts and Codex Accounts
Use multiple Claude Code and Codex accounts on one computer without logging out. Isolate credentials, preserve native history and track quota per account.
Read guide
20 Claude Code Tips & Tricks to Boost Your Productivity
Boost your Claude Code productivity with 20 tips and tricks: CLAUDE.md files, keyboard shortcuts, parallel sessions, history management, MCP servers, and advanced workflows. Updated 2026.
Read guide
Claude Code YOLO Mode: What It Is, the Risks, and How to Enable It Safely
Learn what --dangerously-skip-permissions does, the risks of Claude Code YOLO mode, and safer Auto Mode and Turbo Mode setups with practical guardrails.
Read guide
Claude Code vs Cursor vs Codex CLI: AI Coding Tools Compared
One is an IDE, two are terminal agents, and the price gap is real. What each tool does best, where each one falls short, and how to pick the right one in 2026.
Read guide
Best MCP Servers for Claude Code: Top 10 Integrations You Need
The best MCP servers for Claude Code: GitHub, Notion, Slack, Supabase, Playwright, PostgreSQL and more. What each one does, why you need it, and how to set it up.
Read guide
Claude Code Agent Swarm: Run Multiple Claude Agents in Parallel
A Claude agent swarm runs several independent Claude Code sessions at once. How it differs from subagents, the 3 ways to set one up, and how to keep it from turning into merge chaos.
Read guide
Claude Code Skills Marketplace: Browse and Install Agent Skills
Browse and install skills for ten coding agents, including Claude Code, Codex CLI, Muse Code, Pi and Devin CLI. Share workflows from one workspace.
Read guide
Claude Code Task Management: A Kanban Board Your AI Agents Update
A built-in kanban board for Claude Code task management. Create tasks, assign them to terminals, and let your AI agents read and update the board over MCP.
Read guide
Claude Code Project Switcher: Jump Between Projects in One Click
A Claude Code project switcher that opens any repo in one click. Save navbar shortcuts with their own colour, icon and presets like resume or Turbo mode.
Read guide
AI Commit Messages: Generate a Good Commit From Your Claude Code Diff
Generate a clear commit message with AI from your staged diff, then stage, commit and push without leaving CodeAgentSwarm. A built-in Git Manager for Claude Code.
Read guide
Claude Code Agent Teams vs CodeAgentSwarm: What Is the Difference?
Compare Claude Code teams of independent sessions with CodeAgentSwarm: context, coordination, providers and how to enable Agent Teams.
Read guide
Claude Code GUI: A Visual Desktop App for Your AI Coding CLIs
A Claude Code GUI is a graphical desktop app on top of the CLI. CodeAgentSwarm gives Claude Code a visual workspace, task board, diffs and notifications.
Read guide
Can You Run Multiple Claude Chats at Once?
Yes. Run several Claude chats in browser tabs or Claude Code sessions in parallel. Compare web chats, terminals, Projects, and shared-file risks.
Read guide
How to Run Claude Code on Windows: Native Install or WSL
Claude Code runs natively on Windows, no WSL required. One-line PowerShell install, requirements, WSL 2 setup, troubleshooting and the best way to run multiple sessions on Windows.
Read guide
The Best Claude Code Setup for Windows: Terminal, Shell and Workflow
The Windows setup that makes Claude Code actually pleasant: Windows Terminal, Git Bash as shell, auto-updates, when to add WSL 2, and how to scale to multiple parallel sessions.
Read guide
How to Install Claude Code on Linux (Ubuntu, Debian, Fedora)
Install Claude Code on Linux in one line, or from the official apt, dnf and apk repositories. Requirements, login over SSH, common errors and how to run several Claude Code sessions at once.
Read guide
How to Share Your Skills Between Claude Code, Codex and Antigravity
Your Claude Code skills only live in Claude. Here is how to copy the same SKILL.md skills to Codex and Antigravity in one click with CodeAgentSwarm, instead of moving folders by hand.
Read guide
Claude Code Dashboard: Manage Every Session From One Window
A Claude Code dashboard lets you monitor and manage every session from one window: per-terminal status, a kanban task board, notifications and history.
Read guide
Claude Code Pricing: Every Plan, and Which One You Need for Parallel Agents
Pro $20, Max 5x $100, Max 20x $200, Team seats, API tokens. Plus what the official page skips: which plan survives 3 or 4 Claude Code agents at once.
Read guide
Codex CLI
Codex Agent Swarm: Run Multiple Codex CLI Agents in Parallel
A codex agent swarm runs several Codex CLI agents in parallel. Learn 3 ways to do it: terminal tabs, tmux, and CodeAgentSwarm, with full-auto and supervision.
Read guide
How to Run Multiple Codex CLI Sessions at Once
Run multiple Codex CLI sessions at once. 3 practical methods: terminal tabs, tmux, and CodeAgentSwarm, plus how to run Codex and Claude Code together.
Read guide
Codex CLI Auto Mode and YOLO: Commands, Approvals and Sandbox
Run Codex CLI in Auto Mode with workspace permissions. Compare auto-approve, --yolo and the older --full-auto flag, with commands and sandbox limits.
Read guide
Codex CLI vs Cursor: Terminal Agent vs AI IDE Compared
Codex CLI vs Cursor compared honestly. One is a terminal coding agent, the other an AI-first IDE. See how they differ, where each shines, and when to use both. 2026.
Read guide
How to Find and Resume Your Codex CLI Conversation History
How to find your Codex CLI conversation history, resume a past session, and search every Codex conversation by keyword. Native Codex resume plus CodeAgentSwarm searchable history across all your agents.
Read guide
How to Run Codex CLI on Windows: Native Install or WSL
Run OpenAI Codex CLI on Windows. Install it with npm in PowerShell or run it inside WSL, fix the common Windows gotchas, and learn the cleanest way to run several Codex sessions in parallel.
Read guide
How to Install and Run Codex CLI on Linux
Install OpenAI Codex CLI on Linux with the official script or npm, sign in over SSH with device auth, fix the usual PATH and npm errors, and run several Codex sessions in parallel.
Read guide
Codex GUI: A Visual Desktop App for OpenAI Codex CLI
A Codex GUI is a graphical desktop app on top of OpenAI Codex CLI. CodeAgentSwarm gives Codex a visual workspace, task board, diffs and notifications.
Read guide
Codex CLI Pricing: Every Plan, and Which One You Need for Parallel Agents
Free tier, Go $8, Plus $20, Pro $100 or $200, Business seats, credits. Plus what OpenAI skips: which plan survives 3 or 4 Codex sessions at once.
Read guide
Antigravity CLI
How to Use Antigravity CLI (agy): Install, Commands, and Multiple Sessions
agy install in one line, plus the essential agy commands: the complete guide to Antigravity CLI, Google's successor to Gemini CLI, and parallel sessions.
Read guide
How to Run Multiple Antigravity CLI Sessions in Parallel
Learn how to run multiple Antigravity CLI sessions in parallel. Three methods compared: terminal tabs, tmux, and CodeAgentSwarm. Step by step, with pros and cons.
Read guide
Antigravity CLI vs Gemini CLI: What Changed and How to Migrate
Gemini CLI was retired and replaced by Antigravity CLI. See what changed between the two, whether you have to migrate, and the exact steps to move over while keeping your config.
Read guide
How to Install Antigravity CLI on Linux (Ubuntu, Debian, Fedora)
Install Antigravity CLI on Linux in one line. Requirements, signing in over SSH or with an API key, common Linux errors like a locked keyring, and how to run several agy sessions at once.
Read guide
Antigravity Plans and Pricing: Free Tier, Rate Limits and Parallel Agents
Antigravity pricing: free tier, Plus around $8, Pro $20, Ultra 5x $100, 20x $200. Plus what several agy sessions at once do to the weekly limit.
Read guide
Antigravity YOLO Mode: How --dangerously-skip-permissions Really Works
How to run agy with --dangerously-skip-permissions, what the flag actually skips, when YOLO mode is worth it, and how to contain the blast radius.
Read guide
Antigravity CLI Conversation History: Where It Lives and How to Resume
How Antigravity CLI stores conversations, why they live under ~/.gemini, how to resume with agy -c and --conversation, and why the transcript does not know which project it belongs to.
Read guide
Antigravity CLI on Windows: Native or WSL, and the Gotchas
Running Antigravity CLI on Windows: native versus WSL, where ~/.gemini actually lives, the UNC path trap, why the quota display can lie, and Windows on ARM.
Read guide
Antigravity Agent Swarm: Running Several agy Sessions in Parallel
How to run several Antigravity CLI agents at once: worktree isolation, why approval mode does not scale, splitting work across agents, and keeping every session supervisable.
Read guide
OpenCode
How to Install OpenCode on Linux (Ubuntu, Debian, Fedora, Arch)
Install OpenCode on Linux with the official one-line script, npm, Homebrew or pacman. Requirements, connecting a provider over SSH, clipboard and PATH fixes, and running several OpenCode sessions at once.
Read guide
OpenCode Agent Swarm: Run Multiple opencode Agents in Parallel
An opencode agent swarm runs several opencode agents in parallel. Learn 3 ways: terminal tabs, tmux, and CodeAgentSwarm, with each agent on any model provider.
Read guide
How to Run Multiple OpenCode Sessions at Once
Run multiple opencode sessions at once. 3 practical methods: terminal tabs, tmux, and CodeAgentSwarm, plus how to run opencode alongside Claude Code or Codex.
Read guide
OpenCode vs Cursor: Open Source Terminal Agent vs AI IDE
OpenCode vs Cursor compared honestly. One is an open-source terminal agent that works with any model provider, the other an AI-first IDE. See how they differ. 2026.
Read guide
How to Run OpenCode on Windows: Native Install or WSL
Run opencode on Windows. Install it with npm in PowerShell or run it inside WSL, fix the common Windows gotchas, and learn how to run several opencode sessions in parallel.
Read guide
OpenCode YOLO Mode: Does opencode Have One? Permissions Explained
How to bypass permissions in opencode: there is no YOLO flag. Autonomy is config-driven in opencode.json. Run opencode full auto without flying blind.
Read guide
How to Find and Resume Your OpenCode Conversation History
opencode resume session in two commands: --continue for the last one, --session for a specific one. Plus keyword search across every opencode conversation.
Read guide
OpenCode Plans and Pricing: No Subscription, and What Parallel Agents Cost
OpenCode CLI is free and open source, Zen is pay per token with no monthly fee. Plus what running several agents at once does to your bill.
Read guide
Kimi Code
How to Install Kimi Code on Linux (Ubuntu, Debian, Fedora)
Install Kimi Code on Linux with the official one-line script or npm. Requirements, x64 and ARM64 support, login over SSH, the Alpine catch, common errors and how to run several Kimi Code sessions at once.
Read guide
How to Use Kimi K3 with Claude Code: Full Setup Guide
Run Moonshot Kimi K3 inside Claude Code through the Anthropic-compatible endpoint. Both setups, the env vars that matter, and the limits nobody tells you about.
Read guide
How to Use Kimi Code: Install, Login, Commands, and the kimi-cli Trap
Kimi Code CLI explained: install Moonshot AI's terminal coding agent, log in, learn the commands that matter, and avoid installing the legacy kimi-cli by mistake.
Read guide
Kimi Code Agent Swarm: Run Multiple Kimi Agents in Parallel
A Kimi Code agent swarm runs several kimi sessions in parallel. Learn 3 ways: terminal tabs, tmux and CodeAgentSwarm, plus how shared usage limits behave.
Read guide
How to Run Multiple Kimi Code Sessions at Once
Run multiple Kimi Code sessions at once. Learn how --continue and /title work, 3 practical methods, and what parallel agents do to shared limits.
Read guide
Kimi Code vs Claude Code: Which Terminal Agent Should You Use?
Kimi Code vs Claude Code compared on model, pricing, config, hooks, MCP and maturity. Where each one wins, what actually transfers, and why you may want both.
Read guide
Kimi Code Pricing: Every Plan, the Limits, and What Parallel Agents Cost
Andante ¥49, Moderato ¥99, Allegretto ¥199, Allegro ¥699, plus API per token. And what the official table skips: what happens with several agents at once.
Read guide
Kimi Code YOLO Mode: The --yolo Flag, Auto Mode and Safe Use
Kimi Code has a real YOLO mode: kimi --yolo auto-approves everything. What the flag does, how --auto and --plan differ, and how to run it without wrecking a repo.
Read guide
Kimi Code Conversation History: Find, Name and Resume Sessions
Where Kimi Code stores your sessions on disk, how to resume the last one with kimi --continue or any one with --session, and how to search your whole history.
Read guide
How to Run Kimi Code on Windows: Installer, Git Bash, and the Rough Edges
Run Kimi Code on Windows: the PowerShell installer, the Git for Windows requirement, the ANSI rendering bug you should know about, and WSL as the reliable fallback.
Read guide
Grok Build
How to Install Grok Build on Linux (Ubuntu, Debian, Fedora)
Install xAI's Grok Build CLI on Linux in one line. Requirements, first login, signing in over SSH with a device code, common errors and how to run several Grok Build sessions at once.
Read guide
How to Use Grok Build: Install, Login and Core Commands
Grok Build is xAI's terminal coding agent. Install the grok command, sign in with an xAI account, learn the key commands and supervise several sessions.
Read guide
Grok Build vs Claude Code: Honest Comparison
Grok Build (xAI grok CLI) vs Claude Code compared honestly: models, auth, permissions, subagents, Windows, and when to run both in CodeAgentSwarm.
Read guide
Grok Build Agent Swarm: Run Multiple grok Terminals
Run several Grok Build (xAI) sessions in parallel. How native subagents differ from a multi-terminal swarm, and how tabs, tmux and CodeAgentSwarm compare.
Read guide
Grok Build Plan Mode: Review and Approve Before It Codes
Grok Build Plan Mode explained: how planning and approval gates work, --permission-mode plan, --no-plan, and when --always-approve is the wrong tool.
Read guide
Grok Build Subagents vs Agent Swarm: Which Parallelism?
Native Grok Build subagents vs a multi-terminal agent swarm: when to use --no-subagents, worktrees, and CodeAgentSwarm supervised terminals.
Read guide
Grok Build Pricing and Access: Plans, API Cost and Parallel Agents
Grok Build is free to try. SuperGrok $30, SuperGrok Plus $100, API $2 in and $6 out. Plus what several sessions at once do to your limits.
Read guide
How to Run Grok Build on Windows
Install Grok Build on Windows with PowerShell or Git Bash, PATH notes for %USERPROFILE%\.grok\bin, WSL tips, and CodeAgentSwarm on Windows.
Read guide
Use Grok Build Coming from Claude Code
Moving from Claude Code to Grok Build: AGENTS.md, skills paths (~/.grok/skills), hooks/MCP differences, and running both in CodeAgentSwarm.
Read guide
Grok Build vs Cursor: CLI Agent or AI IDE?
Grok Build (xAI terminal agent) vs Cursor IDE: different products. When you want a coding CLI swarm versus an AI-native editor, and how CodeAgentSwarm fits.
Read guide
Grok Build Headless Mode for Scripts and CI
Run Grok Build non-interactively with grok -p, output formats plain/json/streaming-json, XAI_API_KEY auth, and agent mode for automation.
Read guide
Grok Build Conversation History: Find and Resume Sessions
Find and resume Grok Build sessions with grok --continue, --resume, grok sessions list/search, export transcripts, and CodeAgentSwarm searchable history.
Read guide
Cursor Agent
How to Install Cursor CLI (Cursor Agent) on Linux
Install Cursor CLI (Cursor Agent) on Linux with the official one-line installer. Requirements, PATH setup, login over SSH, common errors and how to run several Cursor Agent sessions at once.
Read guide
Cursor CLI Guide: Install, Agent Mode and ACP
Install Cursor CLI, sign in and run cursor-agent through ACP. Use Agent, Plan or Ask mode with streaming, permissions, MCP, Skills and images.
Read guide
Cursor CLI Conversation History: Find and Resume Sessions
Find Cursor CLI conversation history, check ACP resume support and reopen compatible Cursor Agent sessions in CodeAgentSwarm.
Read guide
How to Run a Cursor Agent Swarm
Run multiple Cursor Agent CLI sessions in CodeAgentSwarm. Split tasks, use Agent, Plan or Ask, supervise permissions and keep Cursor MCP settings.
Read guide
Cursor CLI vs Claude Code: Which Fits Your Workflow?
Compare Cursor CLI, Cursor IDE and Claude Code by interface, account, project setup and multi-agent workflow. Choose based on how your team works.
Read guide
How to Install and Run Cursor CLI on Windows
Install Cursor Agent CLI natively on Windows with the official PowerShell installer, sign in, avoid the Grok agent conflict and start ACP Chat.
Read guide
Cursor CLI Pricing and Usage: Plans, Usage Pools and Parallel Agents
Cursor CLI pricing: Hobby free, Pro $20, Pro+ $60, Ultra $200, Teams from $40. Plus what several Cursor Agent sessions do to your usage pools.
Read guide
Pi coding agent
Pi coding agent: installation, login and your first task
Learn what Pi coding agent does, install the current npm package, connect a model and resume sessions. Includes Pi support in CodeAgentSwarm 2.4.0.
Read guide
Pi models and subscriptions: ChatGPT, Claude and local providers
Connect Pi to ChatGPT, Claude, APIs or local models. Understand subscription billing, sign in with /login and fix missing providers in the model picker.
Read guide
Pi vs OpenCode: models, tools, permissions and integration
Compare Pi and OpenCode by model access, plan mode, permissions, MCP and integration protocol. Choose a coding workflow without confusing agents with models.
Read guide
Run Pi coding agent on Windows: installation and PowerShell
Install Pi coding agent on Windows, choose Git Bash or its PowerShell tool, fix PATH problems and understand how CodeAgentSwarm with Pi runs commands.
Read guide
How to Install Pi Coding Agent on Linux (Ubuntu, Debian, Fedora)
Install Pi coding agent on Linux with the official installer or npm. Requirements, /login over SSH, common Linux errors and how to run several Pi sessions at once.
Read guide
Devin CLI
SWE-2 benchmarks: results, limits and how to test it
Compare published SWE-2 benchmarks with Fable 5.1 and GPT-6 Astra. Understand the evaluation setup and test a real repository task before choosing a model.
Read guide
Devin CLI: installation, login and your first coding task
Install Devin CLI, sign in and run a first task you can verify. See real CodeAgentSwarm beta screenshots and learn where models, quota and history fit.
Read guide
Devin CLI on Windows: native install and troubleshooting
Set up Devin CLI natively on Windows with PowerShell. Check PATH, sign in and locate its files, with real Windows CodeAgentSwarm beta evidence.
Read guide
Devin Pro pricing, free SWE-2 and CLI usage limits
Check Devin Pro pricing, the free SWE-2 promotion and its conflicting official end dates. Understand model selection, context tokens and daily or weekly quota.
Read guide
Devin CLI MCP and history: connect tools and resume work
Configure Devin CLI MCP by scope, inspect connected tools and resume saved conversations. Includes history and optional MCP in CodeAgentSwarm 2.4.0.
Read guide
How to Install Devin CLI on Linux (Ubuntu, Debian, Fedora)
Install Devin CLI on Linux with the official one-line script. Requirements, where it installs, login on a server over SSH, common errors and how to run several Devin sessions at once.
Read guide
Muse Code
How to Use Muse Code: Install, Sign In and Run a Task
Install Muse Code, choose Muse Spark 1.3, sign in and verify your first coding task. Learn project rules, resume commands and CodeAgentSwarm 2.4.0 support.
Read guide
Muse Code on Windows: PowerShell Setup and Troubleshooting
Install Muse Code on Windows with PowerShell, check PATH and sign in. Diagnose shell, sandbox and account issues, and understand current platform limits.
Read guide
Muse Code Models, Pricing and Privacy: What to Check
Compare Muse Code Standard and Contributor models, API token prices and subscriptions. Check training, retention and credentials before sending project code.
Read guide
Muse Code MCP, Skills and Workflows: A Practical Setup
Set up Muse Code MCP and reusable skills, diagnose required-server failures and check workflow availability before designing parallel agent tasks.
Read guide
How to Install Muse Code (Meta) on Linux
Install Meta's Muse Code on Linux with the official one-line installer. Requirements, browser or API key login, SSH servers, common errors and how to run several Muse sessions at once.
Read guide