What problem does it solve?
AI agents can leak API keys, PII, and credentials or fall victim to prompt injection and dangerous commands, and users lack visibility into that traffic. This Skill installs and manages ClawVault, a local HTTPS-inspection proxy that monitors and controls AI agent requests.
Core Features & Use Cases
- Guided installation: Sets up ClawVault in a dedicated virtualenv at ~/.clawvault-env with quick, standard, or advanced modes, plus OpenClaw gateway proxy integration.
- Rule generation and testing: Creates security rules from natural language or scenario templates (customer_service, development, production, finance) and runs built-in detection test suites.
- Health monitoring and lifecycle management: Checks proxy and dashboard status on ports 8765/8766, removes proxy configuration, and fully uninstalls when needed.
- Use Case: A developer worried about leaking AWS credentials through an AI coding agent installs ClawVault, applies the development scenario, and reviews blocked requests in the local dashboard.
Quick Start
Ask the agent to install ClawVault in quick mode with the OpenClaw plugin and then run a health check to confirm the proxy and dashboard are running.