Maintaining an AI-built property finance app: proven Codex testing, release and handoff workflows?

I'm building a small family property-management app using ChatGPT for requirements and Codex desktop on Linux for implementation, with TypeScript, PostgreSQL/Supabase and Vercel. It covers renovation forecasts, quotes, costs/payments, operating income and expenses, and family loans. The app is already in use. We have automated checks, isolated synthetic data, financial reconciliation and controlled releases. The next challenge is making development repeatable while keeping everyday forms, mobile workflows and reporting clear for non-technical family users. I'd value practical experience from people who have maintained an AI-built financial or admin app beyond its prototype: 1. How do you keep requirements, acceptance tests, candidate versions and deployed releases aligned across long Codex threads and workspaces? 2. What browser-testing and release-evidence workflow has reliably caught failures that builds and unit tests missed? 3. Which narrowly scoped GitHub skills, agent packages or UI references have improved your results without adding conflicting instructions or excessive process? 4. What broke after deployment, and what did you change? Links to maintained repositories, runbooks, example acceptance records or fictional-data screen recordings would be especially useful. I'm looking to improve the existing app and workflow rather than rebuild it or add a large agent framework.