OpenClaw: Detailed Introduction

[ikaijua]

OpenClaw: Detailed Introduction

An open-source, self-hosted assistant connecting replaceable models and tools to your devices and messaging channels.

• Audience: Users who want a persistent messaging assistant and can maintain its host, models, and access controls.
• Cost: MIT software without an official paid hosted tier; models, hosting, and third-party services cost separately.

Official account & key contributors on X

• OpenClaw (@openclaw)

Table of Contents

Getting started

1. What it is and who it is for
2. Access and setup: your first session

Usage and configuration

3. A concrete task and how to check it
4. Costs: software, models, and execution
5. Configuration and extensions

Limits and references

6. Permissions, data, and limitations
7. Frequently asked questions
8. Official references

Use the contents to jump to each section. If an in-page link does not respond, open it in a new tab.

[ikaijua]

What it is and who it is for

In one sentence: An open-source, self-hosted assistant connecting replaceable models and tools to your devices and messaging channels.

Who it suits: Users who want a persistent messaging assistant and can maintain its host, models, and access controls.

What you can do:

• Run file and tool tasks on a device or server.
• Reach the assistant through messaging platforms.
• Extend continuing work with skills, plugins, and memory.

Where to start: Follow the website and setup guide. Onboarding configures models, workspace, and the Gateway, the local service coordinating sessions and channels.

Cost: MIT software without an official paid hosted tier; models, hosting, and third-party services cost separately.

[ikaijua]

Access and setup: your first session

Follow the website and setup guide. Onboarding configures models, workspace, and the Gateway, the local service coordinating sessions and channels.

After package installation:

openclaw onboard --install-daemon
openclaw gateway status
openclaw dashboard

Confirm one message in the local Control UI before connecting external channels.

[ikaijua]

A concrete task and how to check it

Begin with public files and a new output file, then add channels or schedules.

Example:

Summarize this week’s items with sources. Mark missing information unresolved, save a new Markdown file, and do not send or delete anything.

Check the artifact and tool records. Diagnose channel connectivity separately from task execution.

[ikaijua]

Costs: software, models, and execution

The Foundation project has no official paid hosted service or token. Verify third-party hosting and model offers separately.

Inference, servers, browsers, and tool APIs can still cost money. Monitor invocation frequency and bills for an always-on assistant.

[ikaijua]

Configuration and extensions

The Gateway connects CLI, Control UI, channels, and device nodes. Models and agent harnesses are replaceable; tools, skills, and plugins add capabilities.

Start with one user and one channel. A team deployment needs deliberate access policies rather than shared unrestricted credentials.

[ikaijua]

Permissions, data, and limitations

Main-session tools can run directly on the host unless sandboxing is configured. Pair unknown senders and treat messages and retrieved content as untrusted.

Local state does not prevent configured models, channels, or integrations from processing task content. Do not expose an unauthenticated public Gateway. Security guide

[ikaijua]

Frequently asked questions

Does it include a free model? Configure and pay for model access separately.

Will it reply with the host off? The Gateway and execution host must be running.

Why do messages arrive but tasks fail? Inspect model access, tools, workspace, and sandbox as well as channel credentials.

[ikaijua]

Official references

• Repository
• Getting started
• Security
• Sandboxing
• Foundation