DeClaw – Local-first security for AI agents

Summary: DeClaw provides real-time protection against sensitive data leaks by detecting PII and blocking prompt injections. It acts as a local security layer between AI agents and the internet, securing external network requests without restricting agent functionality.

What it does

DeClaw inspects network traffic from AI agents, redacting personally identifiable information and preventing prompt injection attacks in real time. It functions as a proxy that filters data locally before it reaches external services.

Who it's for

It is designed for users of AI agents, especially those like OpenClaw, who require secure access to personal files, emails, and terminals without isolating the agent.

Why it matters

DeClaw solves the challenge of balancing AI agent usefulness with data security by enabling safe interaction with sensitive information without fully sandboxing the agent.