Security Threat Model

OpenClaw Threat Model

This is a living threat model for OpenClaw deployments. It identifies the primary threat actors, attack surfaces, and recommended mitigations for AI agent workloads.

Threat Actors

Attack Surfaces

SurfaceRiskMitigation
Messaging channelsDirect prompt injectionDM pairing, input sanitization
Web browsingIndirect prompt injection via web contentContent isolation, separate LLM contexts
File systemUnauthorized file access or modificationPath allowlists, read-only mounts
Command executionArbitrary code executionCommand allowlists, sandboxing
API keysKey exfiltration or abuseSecret redaction, environment isolation

Defense-in-Depth Strategy

No single security control is sufficient. OpenClaw's security model layers multiple defenses:

  1. Perimeter: DM pairing and input sanitization block unauthorized access
  2. Policy: Allowlists and approval gates enforce operational boundaries
  3. Isolation: Sandboxing contains potential damage
  4. Monitoring: Logging and alerting detect anomalies in real-time
  5. Response: Emergency stop and incident response procedures limit blast radius
Prompt Guardian
Protect your AI agent from prompt injection and malicious commands.
Explore Prompt Guardian →