Diffity brings AI code review to browser diffs

Strong release if you believe the hard part of AI coding is not generating code, but reviewing, routing, and closing the loop safely.

• –The browser UI is the wedge: GitHub-like diffs and syntax highlighting make agent output inspectable instead of terminal-only.
• –The review-resolve loop is the key product mechanic; it turns comments into actionable follow-through.
• –Tree browsing and guided tours expand the product from code review into repo exploration and onboarding.
• –Broad support for Claude Code, Cursor, Codex, and other agents is a good distribution strategy because it avoids platform lock-in.
• –Multi-project localhost support is a real usability detail, not a demo-only flourish.