Everything You Need
From security policies to mobile approvals, RCCP has all the features you need to safely deploy AI coding agents in production.
Security Policies
Define custom security policies to control what Claude can do. Block sensitive file access, limit large changes, and enforce code review.
- Path-based protection
- File count limits
- Test failure gates
Mobile Approval
Review and approve Claude's actions from anywhere with the mobile-friendly PWA and real-time updates.
- PWA support
- Real-time updates
- One-tap approval
Real-time Notifications
Stay informed with instant notifications in web dashboard, PWA, and Feishu when Claude needs your attention.
- Web + PWA alerts
- Feishu integration
- Action buttons
Zero Trust Architecture
Your code never leaves your machine. Only events, logs, and approval requests are sent to the cloud.
- Local execution
- Event streaming
- No code upload
Low Latency
Optimized for fast approval loops with WebSocket streaming and low-latency event delivery.
- Fast approval loop
- WebSocket streaming
- Instant sync
Multi-Device Support
Connect multiple machines to a single account. Manage all your development environments from one dashboard.
- Unlimited devices
- Centralized control
- Cross-platform
Git Integration
Automatic PR creation and tracking. See Claude's changes before they're committed.
- GitHub integration
- PR previews
- Branch tracking
Audit Logging
Complete audit trail of all Claude actions and approvals. Export logs for compliance.
- Full history
- Searchable logs
- Export options
How It Works
Get set up in minutes with our simple three-step process.
Install the Agent
Run a single command to install the RCCP agent on your development machine. Works on macOS and Linux.
Configure Policies
Set up security policies from the dashboard. Define what files Claude can access and when approval is required.
Start Coding
Use Claude Code as normal. When it needs permission for sensitive actions, you'll get a notification to approve.