Gateway: OC-001

OpenClaw Guide

OpenClaw is an open-source, local-first Gateway control plane for AI agents. This page covers what it is, how to install it, how ClawHub skills work, and which security controls matter before you trust it with messages, files, and connected accounts.

The ecosystem includes ClawHub (skill registry), plugins, multi-provider model support (OpenAI, Anthropic, OpenRouter, Bedrock, Google, and more), built-in memory with SQLite vector search, and a fast-moving community. If you are comparing agent stacks, also review our AI tools directory, AI risk guide, and beginner AI guide.

Who This Page Is For

Developers, tinkerers, and local-first AI users who want a practical OpenClaw overview before installing or exposing it to real accounts.

What You Will Learn

Architecture, installation flow, skills, threat model, hardening basics, and the official repos/resources worth bookmarking.

Primary Intent

This is an independent OpenClaw guide, not the official docs, focused on setup and security tradeoffs rather than marketing copy.

Status: OC-002

Project Status

Reference snapshot of the OpenClaw ecosystem and install requirements as tracked in March 2026.

v2026.2.14
Latest Release
Node 22+
Runtime Required
18789
Default Port
3 Channels
stable / beta / dev
⚠ Security Warning — Read First

OpenClaw is a high-privilege agent. If misconfigured, it can expose your files, credentials, and connected accounts. Do not expose the Gateway port (18789) to the public internet. Use loopback or Tailscale, require auth, and lock down file permissions (~/.openclaw must be private). Installing third-party skills/plugins is equivalent to running arbitrary code — only install what you have reviewed and trust.

Security contact: security@openclaw.ai · Trust repo: openclaw/trust · CVE-2026-25253 (CVSS 8.8): Control UI token exfiltration — patched in v2026.1.29+

Arch: OC-003

Architecture

How OpenClaw works under the hood.

Gateway

The control plane. WebSocket + HTTP on a single port. Handles sessions, routing, integrations, cron/webhooks, and serves the Control UI. Default mode: loopback only.

Control UI

Web dashboard at http://127.0.0.1:18789. Config editor (form + raw JSON), session viewer, model management. Served by the Gateway.

Channels

WhatsApp, Telegram, Slack, Discord, Signal, iMessage, Teams, Google Chat, WebChat + extensions (Matrix, Zalo, BlueBubbles, voice-call). Anti-fragile across devices.

Skills & Plugins

Skills = SKILL.md + files on ClawHub. Plugins = code modules adding commands/tools + Gateway RPC. Registry: clawhub.ai

Memory & Storage

SQLite with optional sqlite-vec for vector search. Remote embeddings via OpenAI/Gemini/Voyage. Experimental QMD backend (BM25 + vectors + reranking, Markdown as source of truth).

Model Providers

First-class support for OpenAI, Anthropic, OpenRouter, AWS Bedrock, Vercel AI Gateway, Google variants, and multiple Chinese providers. Configurable per-channel.

Spec: OC-004

Skills Directory

Representative OpenClaw skills and install commands. Browse more at clawhub.ai.

Frontend Design

Generate distinctive, production-grade UI with bold aesthetic choices — zero AI slop.

UI + Design
openclaw add frontend-design

Commit Message

Craft meaningful, conventional commit messages that tell the story of your changes.

Git + Workflow
openclaw add commit-message

Security Audit

Scan codebases for OWASP vulns, secrets exposure, and dependency risks.

Security
openclaw add security-audit

API Architect

Design RESTful and GraphQL APIs with proper schemas, validation, and docs.

Backend
openclaw add api-architect

Test Writer

Auto-generate comprehensive test suites — unit, integration, and E2E coverage.

Testing + QA
openclaw add test-writer

DevOps Pipeline

Configure CI/CD workflows, Docker containers, and infrastructure-as-code.

Infra + Deploy
openclaw add devops-pipeline
Spec: OC-005

Prompt Library

Battle-tested prompt templates for daily Claude Code workflows.

01

Deep Code Review

Thorough review focusing on architecture, performance bottlenecks, and security vulnerabilities.

02

Refactor Legacy Code

Systematically modernize legacy codebases while preserving functionality.

03

Generate Documentation

Creates comprehensive docs from code — READMEs, API docs, inline comments, and ADRs.

04

Debug & Diagnose

Systematic debugging workflow tracing issues from symptoms to root cause.

Spec: OC-006

Quick Start

Install OpenClaw and start using it quickly. Requires Node.js 22+. If you want broader context first, see our AI basics page and AI tools directory.

# Recommended: one-line installer (macOS / Linux / WSL2)
curl -fsSL https://openclaw.ai/install.sh | bash

# Windows PowerShell
iwr -useb https://openclaw.ai/install.ps1 | iex

# Or manual install via npm
npm install -g openclaw@latest

# Run onboarding wizard + install daemon
openclaw onboard --install-daemon

# Check the Gateway
openclaw gateway status

# Open the Control UI dashboard
openclaw dashboard

# Add a skill from ClawHub
openclaw add frontend-design
# Update to latest (re-run installer)
curl -fsSL https://openclaw.ai/install.sh | bash -s -- --no-onboard

# Switch update channels
openclaw update --channel stable
openclaw update --channel beta
openclaw update --channel dev

# Run diagnostics
openclaw doctor
Sec: OC-007

Hardening Checklist

10 security controls every OpenClaw user should follow.

1
Keep Gateway on loopback
Use --bind loopback. Reduces remote attack surface.
2
Require authentication
Never run unauthenticated on LAN/Internet. Prevents trivial takeover.
3
Treat Control UI links as sensitive
Avoid untrusted gatewayUrl links. CVE-2026-25253 class risk.
4
Lock file permissions
~/.openclaw/openclaw.json must be 600, ~/.openclaw must be 700.
5
Don't expose Canvas to untrusted users
Canvas can host arbitrary HTML/JS. Keep it local.
6
Run openclaw doctor routinely
Detects common security misconfigurations automatically.
7
Minimise installed skills/plugins
Registry has had malicious skills. Only install what you've reviewed.
8
Isolate execution where feasible
Use sandbox/tool policy to limit blast radius of injected prompts.
9
Rotate API keys on migration
Secrets may persist in state dir. Revoke tokens when uninstalling.
10
Monitor advisories continuously
Security posture changes fast. Watch GitHub releases + advisories.
Links: OC-008

Ecosystem

Official resources and community links.

openclaw/openclaw
Core repository — Gateway, CLI, plugins
ClawHub
Official skill registry — browse, search, install
openclaw/trust
Security & trust resources, advisories
onlycrabs.ai
SOUL.md registry — agent personality files
openclaw-ansible
Hardened installation automation via Ansible
nix-openclaw
Nix packaging for reproducible installs
FAQ: OC-009

OpenClaw FAQ

What is OpenClaw actually for?

OpenClaw is for running a local-first AI agent gateway that can connect models, tools, and chat platforms in one place. Its main appeal is flexibility: users can mix providers, install skills, and keep more control over their environment than in a fully hosted agent product.

Is OpenClaw safe for non-technical users?

Only with care. It is powerful, but it is also high privilege. If you do not understand network exposure, authentication, plugin trust, and file permissions, it is easy to misconfigure. That is why the hardening checklist on this page matters more than the one-line installer.

What should you read after this page?

Start with the official GitHub repo and trust repo, then compare your broader options in our AI tools directory. If you care about risk and governance, continue to our AI dangers page.

View on GitHub →