Security
Security
Hardening an always-on AI agent against real-world threats.
| Document | What It Covers |
|---|---|
| Security Hardening Guide | Comprehensive checklist for locking down an OpenClaw deployment |
| Policy Conformance | Enforce per-agent tool/exec/sandbox posture with the Policy plugin and surface drift |
| Agent-to-Agent Communications | Security analysis of inter-agent messaging with red team test results |
| Why Tailscale SSH | Why Tailscale SSH is fundamentally more secure than traditional SSH for an agent |
| Email Authentication | SPF, DKIM, and DMARC verification to prevent spoofed commands |
| Prompt Injection Defense | Protecting the agent from adversarial inputs in emails and messages |
| Adversarial Review Workflow | Automated tamper detection for agent-pushed commits via GitHub Actions |