MonoClaw

Quickstart

This guide gets you from zero to a working MonoClaw setup that survives real use. Choose a provider, verify a working chat, and know exactly what to do when something breaks.

Since MonoClaw is already pre-installed on your system, you can bypass the installation process and jump straight to choosing a model provider.

The fastest path

Pick the row that matches your goal:

GoalDo this firstThen do this
I just want Mona working on my machinemonoclaw setupRun a real chat and verify it responds
I already know my providermonoclaw modelSave the config, then start chatting
I want a bot or always-on setupmonoclaw gateway setup after CLI worksConnect Telegram, Discord, Slack, or another platform
I want a local or self-hosted modelmonoclaw model → custom endpointVerify the endpoint, model name, and context length
I want multi-provider fallbackmonoclaw model firstAdd routing and fallback only after the base chat works

Rule of thumb: if Mona cannot complete a normal chat, do not add more features yet. Get one clean conversation working first, then layer on gateway, cron, skills, voice, or routing.


1. Choose a Provider

The single most important setup step. Use monoclaw model to walk through the choice interactively:

monoclaw model

Good defaults:

ProviderWhat it isHow to set up
OpenRouterMulti-provider routing across many modelsEnter your API key
AnthropicClaude models directlyAPI key for pay-per-token
OpenAIGPT modelsAPI key
DeepSeekDirect DeepSeek API accessSet DEEPSEEK_API_KEY
Kimi / MoonshotMoonshot-hosted coding and chat modelsSet KIMI_API_KEY
Custom EndpointVLLM, SGLang, Ollama, or any OpenAI-compatible APISet base URL + API key

For most first-time users: choose a provider, accept the defaults unless you know why you're changing them.

Minimum context: 64K tokens

Mona requires a model with at least 64,000 tokens of context. Models with smaller windows cannot maintain enough working memory for multi-step tool-calling workflows and will be rejected at startup. Most hosted models (Claude, GPT, Gemini, Qwen, DeepSeek) meet this easily. If you're running a local model, set its context size to at least 64K (e.g. --ctx-size 65536 for llama.cpp or -c 65536 for Ollama).

How settings are stored

MonoClaw separates secrets from normal config:

  • Secrets and tokens~/.monoclaw/.env
  • Non-secret settings~/.monoclaw/config.yaml

The easiest way to set values correctly is through the CLI:

monoclaw config set model anthropic/claude-opus-4.6
monoclaw config set terminal.backend docker
monoclaw config set OPENROUTER_API_KEY sk-or-...

The right value goes to the right file automatically.


2. Run Your First Chat

monoclaw            # classic CLI
monoclaw --tui      # modern TUI (recommended)

You'll see a welcome banner with your model, available tools, and skills. Use a prompt that's specific and easy to verify:

Summarize this repo in 5 bullets and tell me what the main entrypoint is.
Check my current directory and tell me what looks like the main project file.

What success looks like:

  • The banner shows your chosen model/provider
  • Mona replies without error
  • She can use a tool if needed (terminal, file read, web search)
  • The conversation continues normally for more than one turn

If that works, you're past the hardest part.


3. Verify Sessions Work

Before moving on, make sure resume works:

monoclaw --continue    # Resume the most recent session
monoclaw -c            # Short form

That should bring you back to the session you just had.


4. Try Key Features

Use the terminal

❯ What's my disk usage? Show the top 5 largest directories.

The agent runs terminal commands on your behalf and shows results.

Slash commands

Type / to see an autocomplete dropdown of all commands:

CommandWhat it does
/helpShow all available commands
/toolsList available tools
/modelSwitch models interactively
/personality pirateTry a fun personality
/saveSave the conversation

Multi-line input

Press Alt+Enter, Ctrl+J, or Shift+Enter to add a new line. Shift+Enter requires a terminal that sends it as a distinct sequence (Kitty / foot / WezTerm / Ghostty by default; iTerm2 / Alacritty / VS Code terminal once the Kitty keyboard protocol is enabled). Alt+Enter and Ctrl+J work in every terminal.

Interrupt the agent

If the agent is taking too long, type a new message and press Enter — it interrupts the current task and switches to your new instructions. Ctrl+C also works.


5. Add the Next Layer

Only after the base chat works. Pick what you need:

Bot or shared assistant

monoclaw gateway setup    # Interactive platform configuration

Connect Telegram, Discord, Slack, WhatsApp, Signal, Email, or Home Assistant.

Automation and tools

  • monoclaw tools — tune tool access per platform
  • monoclaw skills — browse and install reusable workflows
  • Cron — only after your bot or CLI setup is stable

Sandboxed terminal

For safety, run the agent in a Docker container or on a remote server:

monoclaw config set terminal.backend docker    # Docker isolation
monoclaw config set terminal.backend ssh       # Remote server

Voice mode

cd ~/.monoclaw/monoclaw-runtime
uv pip install -e ".[voice]"

Then in the CLI: /voice on. Press Ctrl+B to record.

Skills

monoclaw skills search kubernetes
monoclaw skills install openai/skills/k8s

Or use /skills inside a chat session.

MCP servers

# Add to ~/.monoclaw/config.yaml
mcp_servers:
  github:
    command: npx
    args: ["-y", "@modelcontextprotocol/server-github"]
    env:
      GITHUB_PERSONAL_ACCESS_TOKEN: "ghp_xxx"

Editor integration (ACP)

ACP support ships with the standard [all] extras and is pre-installed. Just run:

monoclaw acp

Common Failure Modes

These are the problems that waste the most time:

SymptomLikely causeFix
Mona opens but gives empty or broken repliesProvider auth or model selection is wrongRun monoclaw model again and confirm provider, model, and auth
Custom endpoint "works" but returns garbageWrong base URL, model name, or not actually OpenAI-compatibleVerify the endpoint in a separate client first
Gateway starts but nobody can message itBot token, allowlist, or platform setup is incompleteRe-run monoclaw gateway setup and check monoclaw gateway status
monoclaw --continue can't find old sessionSwitched profiles or session never savedCheck monoclaw sessions list and confirm you're in the right profile
Model unavailable or odd fallback behaviorProvider routing or fallback settings are too aggressiveKeep routing off until the base provider is stable
monoclaw doctor flags config problemsConfig values are missing or staleFix the config, retest a plain chat before adding features

Recovery Toolkit

When something feels off, use this order:

  1. monoclaw doctor
  2. monoclaw model
  3. monoclaw setup
  4. monoclaw sessions list
  5. monoclaw --continue
  6. monoclaw gateway status

That sequence gets you from "broken vibes" back to a known state fast.


Quick Reference

CommandDescription
monoclawStart chatting
monoclaw modelChoose your LLM provider and model
monoclaw toolsConfigure which tools are enabled per platform
monoclaw setupFull setup wizard (configures everything at once)
monoclaw doctorDiagnose issues
monoclaw updateUpdate to latest version
monoclaw gatewayStart the messaging gateway
monoclaw --continueResume last session