OpenClaw CLI Commands: The Complete Cheat Sheet (2026)
OpenClaw has quickly become one of the most talked-about AI agent platforms in 2026, and its command-line interface is where most of the power lives. Whether you are setting up your first agent, managing multi-channel communication workflows, or automating recurring tasks, the CLI is your primary control surface.
This guide covers every OpenClaw CLI command available as of April 2026, organized by category, with full syntax breakdowns, common flags, and practical examples. Bookmark this page --- you will come back to it often.

Table of Contents
- Before You Start: Installation and Prerequisites
- Setup Commands
- Model Management Commands
- Channel Commands
- Skills Commands
- Chat and Messaging Commands
- Cron and Scheduling Commands
- Configuration Commands
- Security Commands
- Status and Logging Commands
- Printable Cheat Sheet Table
- FAQ
Before You Start: Installation and Prerequisites
Before diving into the commands, make sure OpenClaw is installed on your system. The CLI requires Node.js 20+ or can be installed as a standalone binary.
Once installed, you will use the openclaw prefix for every command. The general syntax follows this pattern:
Global flags available on every command:
| Flag | Description |
|---|---|
--help, -h | Show help for any command |
--version, -v | Print the installed version |
--verbose | Enable detailed output |
--json | Output results as JSON |
--quiet, -q | Suppress non-essential output |
--config <path> | Use a custom config file |
For the full official documentation, visit https://docs.openclaw.ai/cli.
Setup Commands
Setup commands handle initial configuration, system health checks, and gateway management. Run these first when installing OpenClaw on a new machine.

openclaw onboard
Runs the interactive onboarding wizard that configures your OpenClaw environment step by step.
Syntax:
Flags:
| Flag | Description |
|---|---|
--skip-models | Skip model provider configuration |
--skip-channels | Skip channel setup |
--non-interactive | Run with defaults (for CI/CD) |
--provider <name> | Pre-select an LLM provider |
Example:
The onboarding wizard walks you through API key entry, default model selection, and optional channel configuration. It writes its results to ~/.openclaw/config.yaml.
openclaw doctor
Diagnoses your OpenClaw installation and reports any issues with dependencies, connectivity, or configuration.
Syntax:
Flags:
| Flag | Description |
|---|---|
--fix | Attempt to auto-fix detected issues |
--check <name> | Run only a specific check |
--list-checks | List all available diagnostic checks |
Example:
Sample output:
openclaw gateway
Manages the local OpenClaw gateway server, which handles inter-process communication, webhook ingress, and skill orchestration.
Syntax:
Actions:
| Action | Description |
|---|---|
start | Start the gateway server |
stop | Stop the gateway server |
restart | Restart the gateway |
status | Show current gateway status |
Flags:
| Flag | Description |
|---|---|
--port <number> | Set the gateway port (default: 7654) |
--daemon, -d | Run in background mode |
--log-level <level> | Set logging verbosity |
Example:
Model Management Commands
Model commands let you configure which large language models OpenClaw uses, check their availability, and manage provider credentials.

openclaw models status
Displays the status of all configured models, including provider, availability, and current default.
Syntax:
Flags:
| Flag | Description |
|---|---|
--provider <name> | Filter by provider |
--ping | Test connectivity to each model |
Example:
Sample output:
openclaw models add
Adds a new model provider or model to your configuration.
Syntax:
Flags:
| Flag | Description |
|---|---|
--key <api-key> | API key for the provider |
--model <name> | Specific model to add |
--default | Set as the default model |
--base-url <url> | Custom API endpoint |
Example:
openclaw models remove
Removes a model provider or specific model from your configuration.
Syntax:
Flags:
| Flag | Description |
|---|---|
--model <name> | Remove a specific model only |
--force | Skip confirmation prompt |
Example:
openclaw models default
Gets or sets the default model used for agent interactions.
Syntax:
Example:
openclaw models test
Runs a quick test prompt against a specified model to verify connectivity and response quality.
Syntax:
Flags:
| Flag | Description |
|---|---|
--prompt <text> | Custom test prompt |
--timeout <seconds> | Request timeout |
Example:
Channel Commands
Channels are OpenClaw's mechanism for multi-platform communication. Each channel represents a messaging platform (email, WhatsApp, Telegram, SMS, Slack) that the agent can use.
openclaw channels list
Lists all configured channels and their current status.
Syntax:
Flags:
| Flag | Description |
|---|---|
--status <active|inactive> | Filter by status |
Example:
Sample output:
openclaw channels add
Adds a new communication channel.
Syntax:
Supported types: email, sms, whatsapp, telegram, slack, discord, webhook
Flags:
| Flag | Description |
|---|---|
--name <name> | Friendly name for the channel |
--provider <name> | Service provider |
--token <token> | Authentication token |
--webhook-url <url> | Webhook endpoint URL |
--default | Set as default for this type |
Example:
For email and SMS channels, Inbounter provides a developer-friendly API that integrates seamlessly with OpenClaw. Its channel setup takes under two minutes.
openclaw channels remove
Removes a configured channel.
Syntax:
Flags:
| Flag | Description |
|---|---|
--force | Skip confirmation |
Example:
openclaw channels test
Sends a test message through a channel to verify it is working.
Syntax:
Flags:
| Flag | Description |
|---|---|
--to <recipient> | Test recipient |
--message <text> | Custom test message |
Example:
Skills Commands
Skills are installable plugins that extend OpenClaw's capabilities. They range from web browsing and file manipulation to domain-specific automation tools.

openclaw skills install
Installs a skill from the OpenClaw skills registry.
Syntax:
Flags:
| Flag | Description |
|---|---|
--version <ver> | Install a specific version |
--global | Install globally (all agents) |
--dev | Install from local directory |
--registry <url> | Custom registry URL |
Example:
openclaw skills uninstall
Removes an installed skill.
Syntax:
Flags:
| Flag | Description |
|---|---|
--purge | Also remove skill data and config |
Example:
openclaw skills list
Lists all installed skills and their status.
Syntax:
Flags:
| Flag | Description |
|---|---|
--enabled | Show only enabled skills |
--outdated | Show skills with available updates |
Example:
Sample output:
openclaw skills search
Searches the skills registry for available skills.
Syntax:
Flags:
| Flag | Description |
|---|---|
--category <cat> | Filter by category |
--sort <field> | Sort by downloads, rating, or date |
--limit <n> | Number of results |
Example:
openclaw skills enable / openclaw skills disable
Toggles a skill on or off without uninstalling it.
Syntax:
Example:
openclaw skills update
Updates one or all skills to their latest versions.
Syntax:
Flags:
| Flag | Description |
|---|---|
--all | Update all installed skills |
Example:
Chat and Messaging Commands
These commands handle direct interaction with the AI agent and message dispatch across channels.
openclaw chat
Opens an interactive chat session with the agent in your terminal.
Syntax:
Flags:
| Flag | Description |
|---|---|
--model <model> | Override the default model |
--system <prompt> | Set a system prompt |
--context <file> | Load context from a file |
--channel <name> | Route responses through a channel |
--history <n> | Load n previous messages |
--no-skills | Disable all skills for this session |
Example:
openclaw send
Sends a single message through a specified channel without starting an interactive session.
Syntax:
Flags:
| Flag | Description |
|---|---|
--to <recipient> | Message recipient |
--subject <text> | Subject line (email only) |
--attach <file> | Attach a file |
--template <name> | Use a message template |
--schedule <time> | Schedule for later |
--ai | Let the agent compose the message |
Example:
openclaw reply
Replies to a specific message in a conversation thread.
Syntax:
Flags:
| Flag | Description |
|---|---|
--ai | Let the agent compose the reply |
--attach <file> | Attach a file |
Example:
Cron and Scheduling Commands
Cron commands let you set up recurring automated tasks that the agent runs on a schedule.

openclaw cron add
Creates a new scheduled task.
Syntax:
Flags:
| Flag | Description |
|---|---|
--name <name> | Friendly name for the job |
--channel <name> | Output channel |
--timezone <tz> | Timezone (default: system) |
--enabled | Enable immediately (default: true) |
--max-retries <n> | Retry count on failure |
Example:
openclaw cron list
Lists all scheduled cron jobs.
Syntax:
Flags:
| Flag | Description |
|---|---|
--active | Show only active jobs |
--sort <field> | Sort by name, next-run, or created |
Example:
Sample output:
openclaw cron remove
Removes a scheduled cron job.
Syntax:
Flags:
| Flag | Description |
|---|---|
--force | Skip confirmation |
Example:
openclaw cron pause / openclaw cron resume
Pauses or resumes a cron job without deleting it.
Syntax:
Example:
openclaw cron run
Manually triggers a cron job immediately, outside its schedule.
Syntax:
Flags:
| Flag | Description |
|---|---|
--dry-run | Show what would happen without executing |
Example:
Configuration Commands
Configuration commands manage OpenClaw's settings, stored in ~/.openclaw/config.yaml.
openclaw config get
Reads a configuration value.
Syntax:
Example:
openclaw config set
Sets a configuration value.
Syntax:
Flags:
| Flag | Description |
|---|---|
--global | Set in global config |
--local | Set in project-local config |
Example:
openclaw config reset
Resets configuration to defaults.
Syntax:
Flags:
| Flag | Description |
|---|---|
--all | Reset entire config to defaults |
--force | Skip confirmation |
Example:
openclaw config edit
Opens the configuration file in your default editor.
Syntax:
Flags:
| Flag | Description |
|---|---|
--editor <name> | Specify editor |
Example:
Security Commands
Security commands help you audit your OpenClaw setup for potential vulnerabilities and manage access controls.
openclaw audit
Runs a security audit of your OpenClaw installation, checking for exposed credentials, insecure configurations, and skill permissions.
Syntax:
Flags:
| Flag | Description |
|---|---|
--fix | Auto-fix issues where possible |
--scope <area> | Audit specific area: config, skills, channels, all |
--report <file> | Export audit report to file |
--strict | Treat warnings as errors |
Example:
Sample output:
openclaw audit permissions
Reviews and manages skill permissions.
Syntax:
Example:
Status and Logging Commands
These commands give you visibility into what OpenClaw is doing, what has happened, and the health of the system.

openclaw status
Shows the current state of all OpenClaw components.
Syntax:
Flags:
| Flag | Description |
|---|---|
--watch, -w | Continuously update (like top) |
--component <name> | Show status of specific component |
Example:
Sample output:
openclaw logs
Displays OpenClaw's operational logs.
Syntax:
Flags:
| Flag | Description |
|---|---|
--follow, -f | Stream logs in real time |
--lines <n>, -n | Number of lines to show |
--level <level> | Filter by log level (debug, info, warn, error) |
--component <name> | Filter by component |
--since <time> | Show logs since a time (e.g., 1h, 2d) |
--search <query> | Search within logs |
Example:
openclaw logs export
Exports logs to a file for sharing or analysis.
Syntax:
Flags:
| Flag | Description |
|---|---|
--format <fmt> | Export format: json, csv, text |
--since <time> | Time range start |
--until <time> | Time range end |
Example:
Additional Utility Commands
A few more commands that are worth knowing.
openclaw upgrade
Upgrades OpenClaw to the latest version.
Syntax:
Flags:
| Flag | Description |
|---|---|
--check | Check for updates without installing |
--pre | Include pre-release versions |
Example:
openclaw reset
Resets parts or all of the OpenClaw installation.
Syntax:
Scopes: config, channels, skills, cron, all
Example:
openclaw completions
Generates shell completion scripts.
Syntax:
Supported shells: bash, zsh, fish, powershell
Example:
Printable Cheat Sheet Table
Here is every command on a single table. Print this out or keep it in a second monitor.

| Category | Command | Description |
|---|---|---|
| Setup | openclaw onboard | Interactive setup wizard |
openclaw doctor | Diagnose installation issues | |
openclaw gateway start|stop|restart|status | Manage the local gateway | |
| Models | openclaw models status | List configured models |
openclaw models add <provider> | Add a model provider | |
openclaw models remove <provider> | Remove a model provider | |
openclaw models default [model] | Get/set default model | |
openclaw models test [model] | Test model connectivity | |
| Channels | openclaw channels list | List communication channels |
openclaw channels add <type> | Add a new channel | |
openclaw channels remove <name> | Remove a channel | |
openclaw channels test <name> | Test a channel | |
| Skills | openclaw skills install <name> | Install a skill |
openclaw skills uninstall <name> | Remove a skill | |
openclaw skills list | List installed skills | |
openclaw skills search <query> | Search skill registry | |
openclaw skills enable|disable <name> | Toggle a skill | |
openclaw skills update [name] | Update skills | |
| Chat | openclaw chat | Interactive chat session |
openclaw send <channel> <msg> | Send a message | |
openclaw reply <id> <msg> | Reply to a message | |
| Cron | openclaw cron add <schedule> <cmd> | Create scheduled task |
openclaw cron list | List cron jobs | |
openclaw cron remove <id> | Remove a cron job | |
openclaw cron pause|resume <id> | Pause/resume a job | |
openclaw cron run <id> | Manually trigger a job | |
| Config | openclaw config get <key> | Read a config value |
openclaw config set <key> <val> | Write a config value | |
openclaw config reset [key] | Reset to default | |
openclaw config edit | Open config in editor | |
| Security | openclaw audit | Run security audit |
openclaw audit permissions | Review skill permissions | |
| Status | openclaw status | System overview |
openclaw logs | View operational logs | |
openclaw logs export <file> | Export logs to file | |
| Utility | openclaw upgrade | Upgrade OpenClaw |
openclaw reset [scope] | Reset installation | |
openclaw completions <shell> | Generate shell completions |
FAQ
How do I update OpenClaw to the latest version?
Run openclaw upgrade. Use openclaw upgrade --check first to see what version is available without installing it.
Where is the OpenClaw configuration file stored?
The global config lives at ~/.openclaw/config.yaml. Project-local configs are stored in .openclaw/config.yaml in your project root.
Can I use multiple LLM providers at the same time?
Yes. Add as many providers as you want with openclaw models add. Each model is identified by provider/model-name. Set your preferred default with openclaw models default.
How do I add email or SMS capabilities?
Use openclaw channels add email or openclaw channels add sms. For a reliable API-first approach, Inbounter provides Email and SMS APIs that are designed for AI agent integration and work seamlessly with OpenClaw's channel system.
Can I run OpenClaw without the gateway?
Some basic commands like openclaw chat work without the gateway, but channels, cron jobs, and skills require the gateway to be running. Start it with openclaw gateway start -d.
How do I debug a failing cron job?
Use openclaw cron run <name> --dry-run to test the job without executing it, then check openclaw logs --search "cron:<name>" for error details.
Is OpenClaw safe to use with API keys?
Run openclaw audit to verify your setup. API keys are encrypted at rest, and the audit command checks for common misconfigurations like exposed credentials or overly permissive file permissions.
How do I contribute a custom skill?
Build your skill following the OpenClaw Skills SDK, test it locally with openclaw skills install ./my-skill --dev, and submit it to the registry via the OpenClaw community portal.
Wrapping Up
This cheat sheet covers every OpenClaw CLI command available as of April 2026. The CLI is the fastest way to configure, automate, and interact with your OpenClaw agent, and knowing these commands will save you significant time as you build more sophisticated workflows.
For the latest updates and new commands, check the official OpenClaw documentation or run openclaw --help at any time.
If you are building AI agents that need reliable email and SMS delivery, take a look at Inbounter's Email and SMS API --- it is built specifically for the programmatic, high-volume messaging that AI agents require.
Related articles: