INSTALL OPENCLAW

Features

Install OpenClaw — Everything Your AI Agent Needs

installopenclaw.io provisions, configures, and maintains your OpenClaw instance so you never have to SSH in for a routine update again. Managed infrastructure, platform connectors, multi-model support, and airtight security — all from a single deployment command.

Managed Hosting

Hands-off server management

Every instance ships with production-grade ops baked in. No Ansible playbooks, no cron jobs, no 3 AM pages.

Zero-downtime rolling updates applied automatically
SSL certificates issued and renewed via Let's Encrypt
Health-check pings every 30 seconds with instant alerts
Nightly encrypted backups with 30-day retention
Resource usage dashboards for CPU, RAM, and disk
Automatic log rotation to prevent disk exhaustion
Firewall rules pre-configured for agent traffic only
One-command rollback to any previous release
Kernel-level process isolation between tenants
Scheduled maintenance windows you control

Integrations

Connect to the platforms your users already live on

Each connector is maintained alongside the core agent — protocol changes and API deprecations are handled before they reach you.

WhatsApp

Stable

12 features

  • Business API media handling
  • Group thread awareness
  • Template message dispatch
  • Read receipt tracking

Telegram

Stable

15 features

  • Inline query responder
  • File and voice handling
  • Channel post automation
  • Keyboard markup support

Discord

Stable

11 features

  • Slash command registration
  • Thread and forum support
  • Embed builder utilities
  • Role-gated permissions

Slack

Beta

14 features

  • App Home surface rendering
  • Workflow step actions
  • Block Kit message composer
  • Threaded conversation context

AI Models

Swap models without changing a line of config

Point your agent at a cloud API or a local inference server. Switching providers is a single environment variable.

Anthropic Claude

Cloud

Strong at sequencing multi-step operations where each action depends on the result of the last. Well-suited for infrastructure orchestration tasks.

OpenAI GPT

Cloud

Versatile general-purpose model effective for content generation, data classification, and structured output formatting.

Ollama

Self-hosted

Run open-weight models on your own hardware. All inference stays on your network — no external API calls leave the machine.

llama.cpp

Self-hosted

Lightweight C++ inference runtime that runs quantized models on modest hardware. Ideal for air-gapped deployments with limited resources.

Developer Tools

Build on top of your agent

Every operational surface is exposed through stable interfaces so your team can automate, extend, and integrate without friction.

SSH & CLI Access

Full shell access to your agent instance. Inspect processes, tail logs, or run ad-hoc commands directly on the host.

REST API

Programmatic control over every agent capability — trigger tasks, read state, and manage configuration through versioned endpoints.

Custom Domains

Point your own domain at the agent dashboard. CNAME records are verified automatically and SSL is provisioned on attachment.

Real-time Log Streaming

Pipe agent output to your terminal or log aggregator as events happen. Supports stdout, stderr, and structured JSON formats.

Outbound Webhooks

Push notifications to any HTTP endpoint when jobs complete, errors surface, or thresholds are breached. Configurable retry policies included.

Secrets Vault

Store API keys, tokens, and credentials in an encrypted vault. Values are injected at runtime and never written to disk in plaintext.

Security

Your server, your perimeter, your keys

We designed the hosting layer so that sensitive data never leaves your infrastructure. Every security control is on by default.

TLS Everywhere

Every connection — dashboard, API, agent-to-model, webhook delivery — is encrypted with TLS 1.3. No exceptions.

Root Credential Retention

You hold the only root keys. Our provisioning system never stores or transmits superuser credentials.

Comprehensive Audit Logs

Every configuration change, login event, and agent action is recorded with actor identity, timestamp, and payload hash.

Network Perimeter Controls

Ingress is restricted to agent-specific ports. Egress can be locked to an allowlist of domains and IP ranges you define.

Encrypted Backups at Rest

Backup archives are AES-256 encrypted before leaving the host. Decryption keys are derived from a passphrase only you possess.

Automated CVE Patching

Security advisories are monitored continuously. Critical patches are staged and applied within hours of public disclosure.

Role-based Access Control

Assign operator, viewer, or admin roles to team members. Permissions cascade through API access, dashboard visibility, and agent control.

Compliance-ready Architecture

Data stays on your server in your jurisdiction. Infrastructure patterns align with GDPR, HIPAA, and SOC 2 control requirements.

Ready to deploy?

Pick a plan, run the install script, and your OpenClaw agent is live in under five minutes.

View plans Sign in to dashboard