Delegate a long data migration with explicit invariants, checkpoints, validation queries, and a reversible cutover.
Source episode · 11:00
How I AI: My GPT-5.5 Review—A 6-Hour Autonomous Task and the Bluetooth Hack No Other Model Could Solvewith Claire Vo
11:00 to 15:25: Claire gives GPT-5.5 Pro a complex data migration, lets it work for roughly six hours, and inspects the completed transformation and checks.
A migration script and run record that move the intended data without losing identity, relationships, or totals.
OpenAI's cloud-based AI software engineering agent that can execute code, run tests, and handle complex multi-file tasks autonomously.
Step by step
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
4 steps
Define the scope of the migration. This approach is best for complex tasks, such as migrating millions of database records stored in various legacy formats with numerous edge cases.
Describe the source problem, target model, affected volume, and why the current data is wrong.
In an environment like Codex, provide a high level, goal oriented prompt. Instruct the agent to perform the migration and also to build a system to test its own work, identify failures, and repair them.
Write and run a dry-run migration with field mappings, invariants, batch checkpoints, validation queries, and rollback steps; do not cut over production.
Allow the model to execute the task autonomously. The process can take several hours with little to no human intervention as the AI creates sub agents and self healing loops to complete the work.
Continue through the approved dry-run, saving progress and reporting any invariant failure immediately.
When the agent's work is complete, review the outcome. For a large migration, expect very few remaining edge cases. Monitor error rates in a tool like Sentry to confirm the impact.
Compare source and target counts, relationships, sampled records, and application behavior; monitor the approved cutover separately.
Turn an idea into a PRD, user stories, and a plan.
After the steps
How to recover when the loop fails and where human judgment helps.
Recover
Keep building
Create context-aware guardrails for your enterprise AI agent using a 'Project' governance layer. As a project owner, define rules, scope tools, and require human-in-the-loop approval for sensitive actions, ensuring safe and effective AI use.
Transform a successful AI chat session into a reusable skill that anyone in your company can use. Package prompts, tools, and code into a standardized workflow, making it easy to share and scale automation across your organization.
Use a natural language AI agent to instantly generate and iterate on custom data dashboards. This workflow queries your company's data securely, allowing non-technical users to perform their own 'last mile' data analysis without filing tickets.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.