Hey Qoder community 
Just added Qoder support to Auxly - a safety layer I built for MCP-compatible AI coding assistants.
Quick backstory: An AI agent almost deleted my production database files a few weeks ago. No approval prompt, just action. That’s when I realized we need better safety infrastructure for AI coding.
What Auxly adds:
Sound alerts when AI asks clarifying questions
Approval workflows for critical changes
One-click undo for any AI action
Token usage tracking per task (monitor AI consumption)
Complete audit trail
Privacy-first architecture
Why this matters for multi-IDE users:
Switch between Qoder, Cursor, Windsurf, or Trae - all your tasks and context come with you. The AI agent remembers everything from previous sessions without re-explaining scope. Perfect for developers who rotate between IDEs when they hit token limits.
For Qoder users specifically:
Works through MCP (Model Context Protocol) integration, so it layers on top of Qoder’s existing AI capabilities without interfering with your workflow.
Tech details:
Built with TypeScript + React
MCP protocol integration
Cross-IDE task synchronization
Developed in Jeddah, Saudi Arabia (Tzamun Arabia IT Co.)
Links:
Website: https://auxly.tzamun.com
Install: https://open-vsx.org/extension/Auxly/auxly-extension
Discord: Auxly
What safety features would be most valuable for your Qoder workflow? Let me know 
