ACE AI SW Startup Training Day 1 Textbook: Codex Requirements Definition Guidebook
← Back to LMS AI Classroom
DAY 1 TEXTBOOK ACE AI SW Startup Training
ACE AI SW Startup Training Textbook Series 01

Codex Client-Based
Requirements Definition Textbook

This textbook covers the practical process in generative AI and Codex Client, from cleaning up unstructured transcripts to sorting user roles, writing functional requirement tables, designing normal/exception scenarios, building open questions and an automated email pipeline.

ACE AI SW Startup Training | Day 1 Textbook Study guide
HOW TO STUDY

How to study this textbook on your own

The goal of this textbook is not to ask AI for a "plausible-looking document." It's to turn facts confirmed in the meeting source into traceable requirements, and to make specs developers can use for building and testing.

Prerequisites: basic file and folder use Suggested time: concepts 60 min Lab 90 min Review 30 min

What you can do after studying

01. Context design

Explain the difference between one-off chat and Workspace-based work.

02. Sorting information

Separate roles, features, rules and undecided items from meeting notes.

03. Writing requirements

Write functional requirements in a checkable format.

04. Scenario design

Design the normal flow together with key exception flows.

05. Checking AI results

Check for source evidence, gaps, contradictions and made-up content.

06. Finishing the final spec

Combine outputs, get human approval, then share.

Suggested study routine

  1. Read each section's concepts and sum up the key word in one sentence in your own words.
  2. Before copying a prompt as-is, check the input files and output conditions.
  3. In example results, separate "sentences with source evidence" from "sentences needing confirmation."
  4. Solve the end-of-chapter problems first, then check the folded model answers.
ACE AI SW Startup Training | Day 1 Textbook Table of Contents
CONTENTS

Table of Contents

💡 How to use this textbook

This textbook includes lab code and prompt text. Use the [🖨️ Save as PDF / Print] button at the top to keep it as a clean A4 book-style PDF.

Chapter 1. Codex Client Overview and Understanding the Workspace Day 1 Textbook
CHAPTER 01

Codex Client Overview and Understanding the Workspace

Key question: what do we give AI, and what do we check?Study time: 35 min

1.1 ChatGPT questions vs Codex Workspace design

In typical generative AI use, people type a one-off prompt into the chat each time and get a short answer. But when building real business software and agent systems, you need more than quick questions and answers: you need 'system design' that keeps the many related files and rules in a project consistent.

Type ChatGPT prompt questions Codex Workspace system design
Kind of work One-off Q&A (single conversation) Continuous project watching and file sync
Keeping Context Depends on each person's prompt skills Automatic caching and mapping of context across files
Output form Simple text answer Detailed markdown (MD) files in a local folder
Outside connections Not possible (just chat) Pipelines connected to outside APIs (Gmail, etc.)
Key idea: Context is not just "memory"

Context is the bundle of material and rules AI uses for its current judgment. Just because a file is in the Workspace doesn't mean it's all understood correctly. Make file names, latest versions, priorities and forbidden rules clear to reduce conflicts.

1.2 The 3-stage pipeline for business requirements

For an AI agent to handle business requirements precisely, beyond simple chat, it must go through this 3-stage pipeline.

1

Context Alignment

The stage where all docs in the project (meeting notes, feature specs, business rules) are synced under one set of rules.

2

Execution Power

Goes beyond writing docs to actually connect with outside APIs and modules (e.g. automatic email).

3

Structured Output

Produces a detailed markdown doc (`requirements-specification.md`) that makes testing and code generation easy.

Concept check 1

Situation: a teammate edited the meeting notes, but AI kept using the old requirements. What should you check first?

  1. Was the latest file saved in the Workspace?
  2. Does a similar old file name cause a conflict?
  3. Is there a rule for which doc takes priority?
Show model answer and explanation

You need to check all three. In particular, first writing a rule like "use the latest file, and ask the user if there's a conflict" reduces the risk of AI picking a version on its own.

Chapter 1. Codex Client Overview and Understanding the Workspace Day 1 Textbook

1.3 Workspace context and multi-file sync

Simply put, a Workspace is a 'local shared planning folder' you give AI. When you make a lab folder (`automation/`) on your desktop and put related files in it, Codex Client treats the whole folder as its working memory (Context).

🔍 Multi-file caching & reasoning structure

• Watching: watches all docs in the folder, like meeting-notes.md and open-questions.md
• Updating together: when one spec changes, updates all linked docs at once without data conflicts
• Consistent reasoning: applies Anti-Hallucination so it doesn't invent content not in the source docs

📌 Why cleanup (step 0) and object sorting (step 1) are needed

The raw meeting transcript (`meeting-notes-raw.md`) mixes in small talk and noise, so developers and designers can't use it directly as a spec. So it's essential to first make a plain-sentence key summary of 15 lines or fewer (`step 0`), then catalog it into user roles and feature objects (`step 1`).

Type Step 0. Clean up the meeting summary Step 1. Sort planning objects
Main purpose Filter back-and-forth chat and noise Catalog cleaned info by type
Action Informal summary in 15 plain sentences or fewer Separate user roles, features, rules and undecided items
Output file Create meeting-notes.md Define 4 kinds of planning component objects

Suggested lab folder structure

automation/
automation/ ├── 00-source/meeting-notes-raw.md # original, never edited ├── 01-working/meeting-notes.md # cleaned working doc ├── 02-spec/open-questions.md # undecided items ├── 02-spec/requirements-specification.md └── AGENTS.md # work rules and don'ts

What a person must always check when using AI

  • Were users, features or policies added that aren't in the source?
  • Are "possible" and "decided" mixed up?
  • Are personal data, recipient addresses and outgoing content correct?
  • Is the same term used with the same meaning across docs?
Chapter 2. 6-Step Lab Guide to Codex-Based Requirements Day 1 Textbook
CHAPTER 02

6-Step Lab Guide to Codex-Based Requirements

Bring: raw meeting notesOutput: 3 Markdown filesStudy time: 70 min

This chapter explains the 6-step lab process from an unstructured customer consultation transcript to the final email sharing pipeline, with the prompt and expected output for each step.

2.1 [Step 0] Clean up a meeting summary (`meeting-notes.md`)

Read the raw transcript (`meeting-notes-raw.md`), remove greetings and small talk, and create clean meeting notes of 15 plain sentences or fewer focused on business requirements.

Prompt template: step 0 cleanup instruction
Read the raw meeting notes in meeting-notes-raw.md, remove unneeded conversation and context between speakers, and write compact, informal summary notes of about 15 lines covering service users, feature requirements, system policies and undecided items. Don't imagine and add anything that isn't in the original.
📥 Input / 📤 Output

Input: meeting-notes-raw.md (raw transcript of about an hour of consultation)
Output: meeting-notes.md (cleaned key planning summary of 15 lines or fewer)

A summary to avoid

"For user convenience, provide easy login and recommendations." Adding features not in the source because they seem like common sense distorts the requirements.

A good summary

"Customers view experts' available times and book a session. The login method was not decided in the meeting." It separates facts from undecided items.

2.2 [Step 1] Sort planning objects from meeting notes

Based on the cleaned `meeting-notes.md`, separate the service's main user roles, features users perform, features the system performs, business rules and undecided items.

Prompt template: step 1 planning element analysis
Read meeting-notes.md and separate the following. 1. Service users (User Roles) 2. Features users perform (User Actions) 3. Features the system performs (System Actions) 4. Business rules (Business Rules) 5. Things not decided yet (Open Questions) Don't add anything not in the meeting notes, and mark unclear items separately.

Example sorting result

CategoryExampleReason
User rolesCustomer, expert, adminWho acts or holds permissions
User featuresChoose session timeDone directly by the user
System featuresSend booking confirmation noticeDone automatically when conditions are met
Business rulesBooking confirmed after payment succeedsConditions and limits of the work
UndecidedCancellation fee rateNeeds approval or more discussion

Step 1 done criteria

Don't put the same item in two categories, and leave a meeting-notes sentence or line number as evidence next to every item. If there's no evidence, move it to "undecided", not "guess".

Chapter 2. 6-Step Lab Guide to Codex-Based Requirements Day 1 Textbook

2.3 [Step 2] Write standard functional requirement sentences

Using the sorted objects, write clear, short sentences that fit the requirements spec table format.

✍️ 4 rules for writing requirement sentences

1. Name the subject: write the subject clearly, like '[The Customer]' or '[The system]'.
2. Clear object: say what data is handled (e.g. expert list, payment method).
3. Same ending: use the same style, like 'must be able to ~' or 'must ~'.
4. Short sentences: put only one feature in each sentence.

Prompt template: step 2 requirements table sentences
Based on the planning elements analyzed in step 1, write a Functional Requirements Table. Include: ID, type (user/system), subject, feature description, business rule Sentence form: "[Subject] must be able to [feature] when [condition]."

A bad requirement

"The system must handle bookings well, quickly and conveniently." Fast, convenient and well can't be measured.

An improved requirement

"When payment is approved, the system must change the booking status to 'confirmed' and send a notice to the customer and expert." The condition, state change and recipients are visible.

Formula for a checkable requirement

[Actor] must [action] on [input/target] when [precondition], and provide [an observable result].
Developers and testers reading the sentence should picture the same screen and data state.

2.4 [Step 3] Design the Happy Path and Edge Cases

Design the basic steps that complete normally when users use the service, and the handling flow when exceptions happen, like a card payment failure or server error.

Flow type Normal flow (Happy Path) Exception flow (Edge Case)
Session booking scenario Choose expert ➔ choose time ➔ payment succeeds ➔ booking confirmation sent Card over limit/error ➔ failure message pop-up ➔ prompt to retry payment
Expert approval scenario Submit documents ➔ admin review ➔ approved ➔ profile goes live Required documents missing ➔ write rejection reason ➔ email asking expert to resubmit

5 questions for finding exception flows

  1. What if required input is missing or in the wrong format?
  2. What if a user without permission tries?
  3. What if two users pick the same resource at the same time?
  4. What if an outside payment, email or IoT device doesn't respond?
  5. What if the network drops or the user cancels mid-task?
Mini lab 2

Turn "the user books a seat" into 1 checkable requirement and 2 exception flows.

Show example answer

Requirement: while logged in, a user must be able to pick an available seat and time and submit a payment request.
Exception 1: if another user pays first and takes the seat, show the latest seat list.
Exception 2: if payment approval fails, don't confirm the booking; show the reason and how to retry.

Chapter 2. 6-Step Lab Guide to Codex-Based Requirements Day 1 Textbook

2.5 [Step 4] Build Open Questions

Organize items not clearly decided in the meeting, like fee policies, refund rules and notification methods, and write a question list doc (`open-questions.md`) you can use right away when interviewing the client.

Prompt for writing open-questions.md
Pull the undecided items from meeting-notes.md and create an open-questions.md file. For each item, include the question, possible options (A/B/C), and how leaving it undecided affects the business.
What makes a good open questionExample
What to decideWhat is the refund rate for same-day booking cancellations?
OptionsA. No refund / B. 50% / C. Operator's choice
ImpactAffects billing logic, terms and customer messages
Decision maker · deadlineService lead · before development starts

2.6 [Step 5] Create the combined requirements spec and share by email

Merge all the markdown docs from steps 0–4 into the final combined spec `requirements-specification.md`, and automatically send a report email to the approver through Codex Client's Gmail API connection.

Instruction for creating the combined doc and sending email automatically
Combine meeting-notes.md, the functional requirements, the normal/exception flows and open-questions.md written so far into a finished requirements-specification.md file. When it's done, write an approval request email with a summary and send it.
📧 Checks for the Gmail lab

Check that the Gmail plugin at the top right of Codex Client is turned on. Before sending, do a final check that the recipient address (`to`), subject (`subject`) and attachment name in the body are correct.

Automate sending, but don't automate approval

Create the email as a draft first; a person checks the recipient, attachments, personal data and undecided markers before sending. During the lab, use your own test address instead of a real customer's.

Suggested table of contents for the final spec

  1. Purpose and scope of the doc
  2. Terms and user roles
  3. Functional requirements and sources
  4. Normal·exception scenarios
  5. Business rules and non-functional requirements
  6. Open questions, decision makers, deadlines
  7. Change history and approval status
Chapter 3. Lab Assignments and Planning Self-Check Day 1 Textbook
CHAPTER 03

Lab Assignments and Planning Self-Check

3.1 [Lab 1] Pet care matching service

An assignment to define requirements for a platform that connects pet owners with pet sitters.

🐾 Key planning mission requirements

• User roles: owner (Customer), pet sitter (Sitter), admin (Admin)
• Key features: view care schedules, request a match, pay a deposit, write a care log
• Undecided: same-day refund rate for cancelled care, whether to require pet sitter identity insurance

3.2 [Lab 2] Self-service study café IoT system

An assignment to define requirements for an IoT system for study room seat booking and door access control via kiosk and mobile app.

🚪 Key planning mission requirements

• User roles: user (Student), store owner (Owner)
• Key features: QR code entry/exit control, remaining seat time alerts, study room light/air conditioning control
• Exception flow: automatic late fee and warning alert when check-out time is exceeded

3.3 4-point planning review self-check list

Review item Detailed self-check criteria Result
1. Roles Are users clearly split into Customer, Expert/Sitter and Admin? [   ] Pass
2. Cause and effect Is the flow from payment success to booking confirmation and notification clear? [   ] Pass
3. Undecided items separated Are pending items like refunds/fees separated out as undecided items? [   ] Pass
4. Fact check Were unrelated features not in the source transcript left out? [   ] Pass
What to submit

Pick one of the two topics and write the following files.

  1. meeting-notes.md: 10–15 fact-based lines
  2. open-questions.md: at least 3 questions, with options and impact
  3. requirements-specification.md: at least 6 functional requirements and at least 4 exception flows

Grading · 100 points

  • Faithful to source, 25 pts: no features or policies without evidence
  • Clarity, 25 pts: actor, condition, action and result are specific
  • Completeness, 25 pts: normal, exception and undecided items all present
  • Traceability, 15 pts: requirement IDs linked to source evidence
  • Doc quality, 10 pts: consistent terms and format, easy to read
Chapter 4. Combined Lab and Model AnswersDay 1 Textbook
CHAPTER 04

Combined lab: turn meeting notes into requirements

Write requirements using only the notes below. The key skill at work isn't filling blanks with imagination; it's separating facts from undecided items.

Sample meeting notes
Students check remaining seats in the app and book in 2-hour blocks. A booking is confirmed only when payment is approved. On entry, students show the app's QR code to the door device. The owner can mark broken seats as not bookable. The payment method and cancellation fee are not decided yet. How to handle entry during a network outage also needs more discussion.
Combined questions
  1. Find all the user roles.
  2. Write 5 functional requirements with IDs.
  3. Write 1 normal flow and 3 exception flows.
  4. Write 3 open questions with priorities.
Show model answer

Roles: student, owner. The door device can be noted separately as an external system that acts.
FR-01 Students must be able to view bookable seats. FR-02 Students must be able to choose a seat in 2-hour blocks. FR-03 The system must confirm only bookings whose payment is approved. FR-04 The system must allow entry when a valid QR code is checked. FR-05 The owner must be able to change a broken seat to not bookable.
Normal flow: view seats → choose time → payment approved → booking confirmed → QR issued → entry.
Exceptions: payment failure, seat taken while choosing, QR check failure. Network outage handling has no policy yet, so don't decide it yourself.
Undecided: P1 entry policy during network outage, P1 payment method, P2 cancellation fee. The first two block building the core flow, so they're high priority.

Day 1 wrap-upKey terms and final check
REVIEW

Key Terms and Pre-Submission Checklist

TermIn plain words
ContextThe range of files, chats and rules AI refers to when making a judgment now
WorkspaceA workspace that manages related material and outputs as one project
Functional RequirementAn observable feature a user or system must perform
Business RuleThe conditions, limits and policies under which a feature works
Happy PathThe main normal flow that reaches the goal without errors
Edge CaseThe handling needed at boundary conditions or in failures
Open QuestionAn unsettled item a responsible person must decide before building
TraceabilityBeing able to trace which source and decision a requirement came from

Final 10-question check

  1. Does every requirement have a unique ID?
  2. Does each sentence have an actor and an observable result?
  3. Does each sentence contain only one feature?
  4. Can you find the source in the meeting notes or a decision record?
  5. Did you separate business rules from features?
  6. Does the core normal flow connect from start to finish?
  7. Did you check permission, input, duplicate and external failure exceptions?
  8. Did you write the impact and decision maker for undecided items?
  9. Did you remove guesses added by AI?
  10. Did a person give final approval before sharing outside?
Day 1 in one sentence

Good AI use isn't about making nice sentences fast; it's about building a structure where facts and decisions can be traced.