Interlock V6: AI agent recovery & verification

Hey Product Hunt! 👋

I'm building INTERLOCK to help address a growing challenge with AI coding agents: what happens when an agent makes a harmful change?

AI agents are becoming increasingly capable of modifying code and project resources. Stopping an agent is important, but developers also need ways to investigate what happened and determine whether affected changes can be recovered safely.

INTERLOCK focuses on AI-agent control, investigation, recovery of supported changes, and verification.

I'm launching to learn from developers who are already using AI coding agents and to understand what they need to trust these tools with more responsibility.

I'd love your feedback:

What worries you most about giving AI agents access to your projects?
What recovery capabilities would make you feel safer using them?
What would you want INTERLOCK to support next?

Thanks for checking out INTERLOCK. I'm excited to learn from this community and keep improving it!