Learn the core control loop: inspect, plan, edit, verify, and iterate with a code agent.
Riley Brown30 minTranscript found
Quick learning frame
Read this before watching.
Coding-agent workflow is the loop of inspect, plan, edit, verify, summarize, and route the next task to the right tool.
This is the fastest practical ramp into using Codex as a daily builder.
Skill you build: The ability to map a knowledge-work or coding task to the right Codex capability — files, memory, plugins, skills, image gen, browser/computer use, or automation — and chain them into a repeatable workflow.
Watch for the shift from claim to mechanism. The learning value is the point where the transcript reveals a repeatable action, tool boundary, context move, review habit, or artifact.
Concept diagram
Where this video fits.
01Inspect
02Plan
03Edit
04Verify
05Review
06Route
Deep lesson
Turn this video into working knowledge.
5,085 cleaned transcript words reviewed across 1,424 timed caption segments.
Thesis
Learn 95% of Codex in 30 minutes teaches a practical codex + claude workflows move: Learn the core control loop: inspect, plan, edit, verify, and iterate with a code agent.
The goal is not to remember the video. The goal is to extract the operating principle, tie it to timestamped evidence, test how far the claim transfers, and make something reusable.
0:32
Files live locally
“cases, so that by the end, you know exactly what it can do and where it fits in your workflow. Let's get started. So, this right here is Codex. This is a clean interface for AI agents that...”
Unlike ChatGPT or Claude where uploads sit in the cloud, everything you give Codex or it creates is stored on your computer, and its agent has full file access. Given a downloads folder of 53 receipt photos, Codex OCRs them, categorizes transactions, and builds an Excel dashboard (total spend, category and payment summaries, monthly trends) you can open in-app or in Finder. Staying organized means working inside a project folder so every output lands there. Point Codex at a folder of messy files you already have (receipts, invoices, screenshots) and ask it to analyze them into a single spreadsheet dashboard, then open the file in your OS to confirm it's local.
9:21
Two kinds of memory
“there's also a different type of memory that you should never really touch, and this is an auto memory feature that Codex kind of keeps updated automatically. And just like the agents.md file, this is all stored in...”
Codex has manual memory and auto memory. Manual memory is the agents.md file — a living document you (or the agent, on request) edit, e.g. 'always use this landing-page format when I don't specify styling.' Auto memory is a separate file Codex maintains itself via a skill; you should observe it but never hand-edit it, since it improves over time and summarizes the tasks you've asked for. Ask Codex to remember one styling or workflow preference and confirm it wrote it to agents.md, then have it open and summarize its auto-memory file so you can see what it has learned about you.
23:44
It drives your screen
“This is very cool. Anything you can open up in the browser, you can ask uh browser use. You can use the @browseruse plugin to test it directly inside Codex. And finally, capability number seven inside Codex is...”
Via the @computeruse and @browseruse plugins, Codex controls your actual computer and browser like a human with mouse and keyboard. It opens the Canva app and builds a five-slide presentation placing one generated image per slide, then turns an index.html into an app and tests it in the browser — clicking the start button, scrolling, taking quizzes, and marking correct answers to verify every part works. Take something you can open in a browser and ask Codex with @browseruse to click through and test that the buttons and navigation actually work, watching where it succeeds or gets stuck.
01
Inspect
Start with this video's job: Learn the core control loop: inspect, plan, edit, verify, and iterate with a code agent. Treat "Inspect" as the outcome you are trying to make visible, not a topic label. Anchor it to 0:32, where the video says: “cases, so that by the end, you know exactly what it can do and where it fits in your workflow. Let's get started. So, this right here is Codex. This is a clean interface for AI agents that...”
02
Plan
Use "Plan" to locate the part of the codex + claude workflows workflow the video is demonstrating. Ask what changes in your real setup if this claim is true. Anchor it to 9:21, where the video says: “there's also a different type of memory that you should never really touch, and this is an auto memory feature that Codex kind of keeps updated automatically. And just like the agents.md file, this is all stored in...”
03
Edit
Turn "Edit" into the reusable artifact for this lesson: A routing matrix for when to use Codex, Claude, browser checks, or manual review. This is where watching becomes something you can inspect and reuse.
04
Verify
Use "Verify" as the application surface. Decide whether the idea touches a browser flow, a local file, a model choice, a source document, a UI, or a review step.
05
Review
Use "Review" to prove the lesson. The evidence should connect back to the video title, transcript anchors, and a concrete output, not a generic best-practice claim.
06
Route
Use "Route" to carry the idea forward: save the prompt, checklist, diagram, or operating rule that would make the next agent run better.
Example
Source-backed work packet
Convert the video into a scoped task that includes the transcript claim, target workflow, acceptance criteria, and proof. The output should be a routing matrix for when to use codex, claude, browser checks, or manual review..
Example
Claim vs. demo brief
Separate what the speaker claims, what the demo actually proves, and what still needs outside verification before you adopt the workflow.
Example
Teach-back module
Transform the lesson into a definition, a mechanism diagram, one misconception, one practice exercise, and a check-for-understanding question.
Do not learn it wrong
Treating the title as the lesson without checking what the transcript actually says.
Letting the prompt drift into generic advice that could apply to any video in the playlist.
Copying the tool setup without identifying the operating principle that transfers to your own stack.
Skipping the artifact, which means the learning never becomes operational or inspectable.
Do not count this as learned until these are true.
01
State the transcript-backed claim in your own words: Learn the core control loop: inspect, plan, edit, verify, and iterate with a code agent.
02
Explain the practical stakes without hype: This is the fastest practical ramp into using Codex as a daily builder.
03
Map the idea onto the Inspect -> Plan -> Edit -> Verify -> Review -> Route sequence and name the weakest link.
04
Produce the artifact and include the evidence that proves it: A routing matrix for when to use Codex, Claude, browser checks, or manual review.
Put it into practice
Give this grounded prompt to Codex or Claude after watching.
You are helping me turn one specific YouTube video into real, durable learning.
Source video:
- Title: Learn 95% of Codex in 30 minutes
- URL: https://www.youtube.com/watch?v=474wZZHoWN4
- Topic: Codex + Claude Workflows
- My current learning frame: Create a Codex project, connect a plugin like Gmail, run a real task (e.g. pull brand-deal emails into a spreadsheet), then say 'turn this into a skill' and schedule it as a weekly automation so the whole workflow repeats itself.
- Why this matters: This is the fastest practical ramp into using Codex as a daily builder.
Transcript anchors from this exact video:
- 0:32 / Evidence 1: "cases, so that by the end, you know exactly what it can do and where it fits in your workflow. Let's get started. So, this right here is Codex. This is a clean interface for AI agents that..."
- 2:07 / Evidence 2: "talk about a little bit later. As we go through the main seven different capabilities of Codex, you'll realize that it is truly a super app that does any coding task or any knowledge work task. Let's not..."
- 6:07 / Evidence 3: "the document. As soon as the agent's done, it's going to put the uh document that it creates in this project. And as you can see here, we have this Word doc that was created, and it's called..."
- 9:21 / Evidence 4: "there's also a different type of memory that you should never really touch, and this is an auto memory feature that Codex kind of keeps updated automatically. And just like the agents.md file, this is all stored in..."
- 13:52 / Evidence 5: "fourth capability that Codex has, which are skills. And you can think of skills on Codex as reusable workflow recipes or SOPs that your agent can use many times over and over again. In Codex, if you go..."
- 23:44 / Evidence 6: "This is very cool. Anything you can open up in the browser, you can ask uh browser use. You can use the @browseruse plugin to test it directly inside Codex. And finally, capability number seven inside Codex is..."
- 27:47 / Evidence 7: "converting that into a skill. Your Codex agent has access to image generation and it is the best image generation model in the world, which is GPT-image-2. We also have browser control and computer use. And this allows..."
Your task:
1. Use the transcript anchors above as the primary source packet. If you add outside context, label it clearly as outside context and keep it secondary.
2. Create a source-check table with columns: timestamp, claim, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable claims from the video. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A routing matrix for when to use Codex, Claude, browser checks, or manual review.
5. Include:
- a plain-English definition of the core idea
- a diagram or structured model using this sequence: Inspect -> Plan -> Edit -> Verify -> Review -> Route
- 3 concrete examples that apply the video idea to real agentic work
- 2 failure modes the video helps prevent
- a checklist I can use the next time I run Codex or Claude
- one practical exercise with a clear done signal
6. Add a "learning transfer" section: what changes in my workflow tomorrow if I actually learned this?
7. Add a "source check" section that cites which transcript anchor supports each major takeaway.
Quality bar:
- Make this specific to "Learn 95% of Codex in 30 minutes", not a generic Codex + Claude Workflows essay.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- If evidence is weak, say what transcript segment or timestamp needs review instead of guessing.
- Finish with a concise artifact I could paste into my learning app.
Misconceptions
What to stop believing.
One agent should do every task.
Different tools have different strengths. Routing is part of the workflow.
More context is always better.
Relevant context helps; stale context causes drift and cost.
Practice studio
Learning only counts when you make something.
01
Transcript evidence map
Separate what the video actually says from what you already believe about the topic.
3 source-backed takeaways with timestamps, confidence, and a transfer note.02
One useful artifact
Apply the video to a real workflow and produce a routing matrix for when to use codex, claude, browser checks, or manual review..
A reusable artifact with a done signal and one verification step.03
Teach-back card
Explain the lesson to someone who has not watched the video yet.
A 90-second explanation, one diagram, one example, and one misconception to avoid.
Recall check
Answer first, then reveal — without rewatching.
How does Codex handle files differently from ChatGPT or Claude?
What is the difference between Codex's manual memory and auto memory?
What do the @computeruse and @browseruse plugins let Codex do?
Source shelf
Use the video as a doorway, then verify with primary sources.
