Claude Code asks for permission 50+ times a session. We auto-approve them.
Zero clicks. Zero interruptions. Just flow.
Every time Claude Code needs to read a file, write code, or run a command, it asks for permission.
Context switching kills productivity. Every prompt breaks your flow state.
Powerful automation that respects your workflow
Detects and approves prompts in ~300ms. You won't even notice it working.
Zero network access. Zero telemetry. All processing happens locally on your machine.
Monitor CMD, PowerShell, Windows Terminal, or Terminal.app simultaneously.
Track detected, approved, and failed prompts. Know exactly what's happening.
Global execution lock, 5-second cooldown, and verification checks prevent mishaps.
MIT licensed. Review the code, fork it, modify it. Full transparency.
Smart automation under the hood
Polls your terminal every 500ms using Windows UI Automation or AppleScript
Regex parser identifies genuine Claude permission patterns
Sends "2 + Enter" to select "allow from this project"
10.15+ (Intel & Apple Silicon)
The app isn't code-signed because signing certificates cost $400/year. The project is open source — review the code yourself. Click "More info" → "Run anyway" to bypass SmartScreen.
100%. The app has zero network access and collects zero data. It only reads terminal text and sends keyboard input (number + Enter). All code is open source for review.
No. It only approves prompts with "allow from this project" options. One-time "Yes/No" prompts are untouched. Plus, you control which terminals are monitored.
Currently designed for Claude Code, but the architecture is extensible. Fork the repo and adapt the regex patterns for your tool.
Close the app from the system tray. Windows: Delete the .exe file. macOS: Drag the app from Applications to Trash. No files installed elsewhere. Clean removal.
Pull requests welcome! Check the contributing guide and open an issue to discuss major changes.
Join developers who've eliminated permission prompt fatigue