The most detailed free FDE + DevOps library: 140+ lessons, 70+ labs and 80 long-form articles, in English and Turkish. Start learning →

Build with AI: From Zero to Your First App · Module 8: Finish, publish and present

Write a README That States Limits and AI Help

The README is the project's handshake: what it is, what it is not, who helped, what remains. This lesson writes one short enough to be read and honest enough to be trusted.

9 min reading

Objectives

  • Write a README that states what the project is and is not
  • Disclose the AI contribution honestly in two sentences
  • State limits: browser-only, per-device data, unfinished items
  • Keep the README short enough to be read fully

Why this matters

Visitors judge unfinished projects by their honesty: a README stating limits earns trust a silent demo never does. Employers, friends and future-you all read the same file. AI disclosure is part of it: hiding the tool's help invites questions later; stating it plainly shows judgment, the skill the whole program taught.

Concepts

Five short sections. What (one line: bakery one-page order site). Use (how to open: address or double-click index.html). Features (three bullets, matching scope). Limits (browser-only, per-device storage, no payment or accounts, known unfinished items named). AI help (two sentences: which tools for what, what you verified yourself). Under one page; longer READMEs are not read.

Honest AI disclosure example. "Drafted with a chat tool (briefs, text variants) and a coding assistant (single-change edits); every fact checked at its source, every fix verified with the course test routine." Names the help, claims the verification. Never "built entirely by AI" (untrue: decisions were yours) and never silent (invites discovery).

Limits as features. "Orders persist in this browser on this device" is not an apology; it is architecture stated plainly. "Multi-device sync out of scope for version one" schedules the future instead of hiding it. Stated limits convert criticism into roadmap.

Worked example

Bakery README: what (one line), use (address plus local open), features (menu, order form with echo, persistent list with delete), limits (browser-only, per-device, no payment/accounts/tracking, gallery deferred), AI help (two sentences as above). Read fully in ninety seconds, trusted immediately, roadmap visible. The project presents itself while its author sleeps.

Expected result: your README in five sections, under one page, with AI help and limits stated.

The common wrong move

README as marketing ("revolutionary AI-powered platform disrupting food"). Hype without evidence destroys the trust the project earned. Plain words, checked facts, named limits: the course voice to the end.

Lab and next step

Lab A24 (paired with lesson 31) finishes with the README review. Next: the capstone build itself, then module checks, the final assessment and the optional completion certificate, all per the course rules.

Quick check

An optional 4-question self-check. Answers never leave your device, are not stored, and never count toward any assessment.

Lesson feedback

No published feedback yet.

Log in and complete the lesson to leave feedback.

Exercise

Write your five-section README under one page with AI disclosure and limits. Read it aloud in under two minutes; cut anything that fails the read.

Pass criteria

Five sections present; AI help in two honest sentences; limits named (browser-only, per-device, out-of-scope items); full read under two minutes.

Log in to track progressFree account: stores only your lesson progress and quiz results.

Next step

Finish the course: project, final, certificate.

Last lesson done. Submit the capstone project, pass the free final assessment, then the optional $15 certificate opens.