Stop fumbling with manual installation and configuration. install9 handles everything — environment detection, dependency management, security hardening, and channel integration — so you can start building with AI agents in minutes.
Automatically detects macOS, Linux, or Windows, installs the right dependencies, and configures your system.
Gateway tokens, file system isolation, and security hardening applied automatically.
Set up Feishu, Slack, Discord and more with guided interactive prompts.
A production-grade installer that handles the hard parts so you can focus on building.
Detects your OS, package manager, shell type, and existing installations. Works with brew, apt, dnf, pacman, apk, winget, choco, and scoop.
Installs Node.js via nvm or fnm, configures PATH, and handles version upgrades seamlessly.
Configures the OpenClaw gateway as a system service (launchd on macOS, systemd on Linux, Scheduled Task on Windows) with secure token management.
Step-by-step guided setup for Feishu/Lark with App ID, Secret, and WebSocket configuration.
First-class support for OpenClaw with an extensible architecture for future platforms.
Multi-channel AI agent platform with 38 channel plugins and 55 built-in skills. install9 handles the full lifecycle — from first install through gateway configuration to channel integration.
Fully supportedclaw-guard audits what OpenClaw exposes on your host system — credentials, ports, processes, and more.
Existing tools check if OpenClaw is configured correctly. claw-guard asks a different question: now that an AI agent has system permissions, what sensitive resources might it access? It audits your host from the outside in.
Checks if ~/.ssh, ~/.aws, ~/.kube and other credential directories are accessible by OpenClaw processes.
Detects wildcard listeners on 0.0.0.0, unexpected outbound connections, and exposed service ports.
Verifies OpenClaw is not running as root and flags dangerous sub-agent flags like --yolo.
Checks for Docker socket exposure and excessive volume mounts that could compromise the host.
Scans shell history, logs, and config files for accidentally exposed API keys and secrets.
Ensures sandbox mode is enabled and container isolation has no escape vectors.