Public deployment should be a separate approval decision, not the default definition of success. This distinction can affect provider fit: CodeWizardsHQ’s ACM document describes a live, publicly accessible project as a defining feature. Families wanting a private prototype should establish whether that expectation is negotiable before enrollment. (codewizardshq.com)
Explore the comparison in this briefing
Release choice
Local demonstration
- When we would use it
- The goal is learning and explanation
- Approval conditions
- Cleared data, no exposed credentials and a reproducible demo
Compare each option against the same criteria. The complete table and qualifications remain in the briefing.
Our suggested release comparison:
| Release choice | When we would use it | Approval conditions |
|---|---|---|
| Local demonstration | The goal is learning and explanation | Cleared data, no exposed credentials and a reproducible demo |
| Restricted demonstration | A few invited reviewers need to try it | Access control, suitable logs, spending controls and a named adult owner |
| Public application | Public use is genuinely part of the learning goal | Separate security, privacy, cost and operational review; clear responsibility after mentorship ends |
For API-based projects, OpenAI’s guidance warns against placing secret keys in browser/mobile code or source repositories. Its production guidance distinguishes spend alerts from hard spending controls. Ask the mentor to demonstrate the actual arrangement rather than merely saying the project has a budget. (help.openai.com)
Our default brief excludes confidential employer material, children’s personal records and high-stakes health or financial advice applications. Prefer invented or appropriately cleared data and a bounded demonstration until the family has separately approved a more demanding use case.
Sources cited on this page
- Advanced Coaching and Mentoring Program A pers ↗codewizardshq.com
- Best Practices for API Key Safety | OpenAI Help Center ↗help.openai.com