Getting Started

Everything you need to install GuardClaw and start protecting your AI agents.

Installation

macOS and Linux

brew install guardclaw/tap/guardclaw

Verify the installation:

guardclaw version

Docker and container-based deployments are also supported. Run the setup wizard after install for details.

Setup

Run the setup wizard. It detects your platform and configures GuardClaw automatically.

guardclaw init

Once configured, restart your AI tool. All 7 defense layers activate immediately with no extra configuration.

Verify everything is working:

guardclaw doctor

Usage

Wrap any command to run it through GuardClaw's defense layers:

guardclaw run -- <your-command>

Check the current configuration and active agents:

guardclaw status

Supported Platforms

Claude Desktop

Auto-configures on setup.

Cursor

Auto-configures on setup.

Claude Code

Auto-configures on setup.

Docker / Kubernetes / Cloud Run

Container support for production environments.

What GuardClaw Does

GuardClaw enforces 7 layers of defense between your AI agents and the actions they take. Every tool call, file operation, and network request passes through a deterministic policy engine before it executes.

No LLM in the security path. Decisions are made by rules, not predictions. Your data stays in your infrastructure.

1,560+ detection patterns ship out of the box across 11 attack categories.

Cloud Dashboard

The optional cloud dashboard provides audit history, agent monitoring, and team management. Connect from the CLI after install:

guardclaw connect --api-key <your-key> --workspace <your-workspace>

Sign in to the dashboard to get your API key and workspace ID. Full documentation, API reference, and configuration guides are available there.

The cloud dashboard is optional. GuardClaw works fully offline with no external dependencies.

Troubleshooting

GuardClaw not appearing in your AI tool: Make sure the binary is in your PATH (which guardclaw), then restart your AI tool completely. Some tools require a full quit and relaunch.

Safe operations being denied: The CLI includes configuration options to adjust enforcement for local development. Run the health check for specific guidance.

Need help? Contact us at security@takeinterest.ai.