If you don’t run Pi locally you’re falling behind…
David Ondrej's course on Pi agent (pi.dev) argues Pi is a minimal, customizable harness — four tools and a ~1,000-token system prompt across 15+ providers — that you adapt to your workflow rather than the reverse. He covers install, context via markdown files, the four ways to extend Pi (agents.md, prompt templates, skills, extensions), and his daily Pi-plus-Codex orchestration workflow inside the CMUX terminal.
David Ondrej47 minTranscript found
Quick learning frame
Read this before watching.
Coding-agent workflow is the loop of inspect, route, plan, edit, verify, summarize, and decide what should be automated next.
New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Skill you build: The ability to configure and progressively extend a minimal Pi agent — and orchestrate it to drive other agents — so it grows more powerful with your own skills, templates, and extensions.
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 context
02Route tool
03Plan work
04Edit safely
05Verify behavior
06Report next step
Deep lesson
Turn this video into working knowledge.
10,022 cleaned transcript words reviewed across 2,718 timed caption segments.
Thesis
If you don’t run Pi locally you’re falling behind… teaches a practical coding-agent workflow move: David Ondrej's course on Pi agent (pi.dev) argues Pi is a minimal, customizable harness — four tools and a ~1,000-token system prompt across 15+ providers — that you adapt to your workflow rather than the reverse. He covers install, context via markdown files, the four ways to extend Pi (agents.md, prompt templates, skills, extensions), and his daily Pi-plus-Codex orchestration workflow inside the CMUX terminal.
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.
0:14
Harness, not product
“even more than cloth code and CEX. Also, it's one of the fastest growing repositories in all of GitHub, meaning soon enough, I think Pi will become the most popular AI agent in the world. Right now, we're...”
Claude Code and Codex are mass-market products — opinionated, bloated, full of guardrails for millions of users. Pi is a harness: the most minimal agent, four tools, a tiny ~1,000-token system prompt (10-15x smaller, tokens you pay for every use), and support for 15+ providers and thousands of models. Its homepage line captures it: adapt Pi to your workflow, not the other way around. Install is a one-liner curl from pi.dev. Install Pi with the pi.dev one-liner, connect a provider (e.g. OpenRouter with a $5-10 topped-up key), and run it to confirm you can swap models with /model.
13:39
Four ways to extend
“remember every single message put it into agents.mmd file. The second thing is a bit more advanced and that is prompt templates. So this is not just simple prompts. This is more like slash commands. So if you...”
The must-learn secret is the four ways to improve Pi, each more powerful than the last: agents.md (always-on context Pi remembers every message), prompt templates (repeatable slash-commands like /review), skills (auto-loaded when relevant, like Claude skills), and extensions (real TypeScript code acting as hooks, hardest to build but most capable — the web-access extension using Exa is one). After config changes, run /reload to apply them immediately. Have Pi create a /review prompt template in your global .pi/prompts folder, run /reload, and confirm the slash command loads the full prompt.
32:44
Pi drives Codex
“You can use any model. If you prefer cloud models, you can use that. And you're fully in control. And that agent is driving Codex, which is right now the most powerful coding harness. Plus, you can save...”
Pi has no sub-agents by design; creator Mario Zechner recommends spawning multiple Pi instances in parallel via tmux or CMUX for full transparency and control. The standout workflow is Pi as orchestrator driving Codex CLI as the coder: you tell Pi your goal in plain English and it launches, reads, polls (sleep then check), and manages Codex panes in CMUX. This also saves cost — run Pi on an OpenRouter key while Codex runs on a ChatGPT subscription doing ~90% of the tokens. In CMUX, ask Pi to launch a couple of Codex CLI panes and manage them toward one goal you state in plain English, watching how Pi polls and steers them.
01
Inspect context
Start with this video's job: David Ondrej's course on Pi agent (pi.dev) argues Pi is a minimal, customizable harness — four tools and a ~1,000-token system prompt across 15+ providers — that you adapt to your workflow rather than the reverse. He covers install, context via markdown files, the four ways to extend Pi (agents.md, prompt templates, skills, extensions), and his daily Pi-plus-Codex orchestration workflow inside the CMUX terminal. Treat "Inspect context" as the outcome you are trying to make visible, not a topic label. Anchor it to 0:14, where the video says: “even more than cloth code and CEX. Also, it's one of the fastest growing repositories in all of GitHub, meaning soon enough, I think Pi will become the most popular AI agent in the world. Right now, we're...”
02
Route tool
Use "Route tool" to locate the part of the coding-agent workflow mechanism the video is demonstrating. Ask what changes in your real setup if this claim is true. Anchor it to 13:39, where the video says: “remember every single message put it into agents.mmd file. The second thing is a bit more advanced and that is prompt templates. So this is not just simple prompts. This is more like slash commands. So if you...”
03
Plan work
Turn "Plan work" into the reusable artifact for this lesson: A coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal. This is where watching becomes something you can inspect and reuse.
04
Edit safely
Use "Edit safely" 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
Verify behavior
Use "Verify behavior" 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
Report next step
Use "Report next step" to carry the idea forward: save the prompt, checklist, diagram, or operating rule that would make the next agent run better.
Example
Source-backed artifact packet
Convert the video into a scoped artifact request that includes the transcript claim, mechanism, acceptance criteria, and proof. The output should be a coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal..
Example
Coding-agent workflow proof brief
Separate what the speaker claims, what the demo actually proves, and what still needs outside verification before you adopt the coding-agent workflow pattern.
Example
Teach-back module
Transform the lesson into a definition, a Inspect context -> Route tool -> Plan work -> Edit safely -> Verify behavior -> Report next step 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.
choosing tools by hype
losing context across agents
letting parallel sessions become invisible
Letting the lesson drift into generic Codex vs Claude comparison.
Letting the lesson drift into feature lists without task routing.
Letting the lesson drift into claims that ignore limits or recovery.
Do not count this as learned until these are true.
01
State the transcript-backed claim in your own words: David Ondrej's course on Pi agent (pi.dev) argues Pi is a minimal, customizable harness — four tools and a ~1,000-token system prompt across 15+ providers — that you adapt to your workflow rather than the reverse. He covers install, context via markdown files, the four ways to extend Pi (agents.md, prompt templates, skills, extensions), and his daily Pi-plus-Codex orchestration workflow inside the CMUX terminal.
02
Explain the practical stakes without hype: New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
03
Map the idea onto the Inspect context -> Route tool -> Plan work -> Edit safely -> Verify behavior -> Report next step sequence and name the weakest link.
04
Produce the artifact and include the evidence that proves it: A coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal.
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: If you don’t run Pi locally you’re falling behind…
- URL: https://www.youtube.com/watch?v=jcUqsNpDDDk
- Topic: Creative Automation
- My current learning frame: Install Pi, add one prompt template and one agents.md preference, then set up a CMUX workspace where Pi orchestrates one or two Codex CLI instances to build a small app from a single plain-English goal.
- Why this matters: New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Transcript anchors from this exact video:
- 0:14 / Evidence 1: "even more than cloth code and CEX. Also, it's one of the fastest growing repositories in all of GitHub, meaning soon enough, I think Pi will become the most popular AI agent in the world. Right now, we're..."
- 6:02 / Evidence 2: "basically just like updating the system prompt without messing up any of the default things that Mario added. The third option is creating agents.mmd file either globally or inside of any folder you're working on with context specific..."
- 10:56 / Evidence 3: "set up a new skill and it can do it. So, a lot of you are limiting the AI by your own limited prompts, by your own low-level prompts because you're not giving it ambitious enough of tasks."
- 13:39 / Evidence 4: "remember every single message put it into agents.mmd file. The second thing is a bit more advanced and that is prompt templates. So this is not just simple prompts. This is more like slash commands. So if you..."
- 16:12 / Evidence 5: "Pi keybinds, extensions, skills, prompts, themes. Just it will reload your entire PI agent. And boom, just like that. If we do SL review, it works now. And as you can see, we have the SL review skill..."
- 32:44 / Evidence 6: "You can use any model. If you prefer cloud models, you can use that. And you're fully in control. And that agent is driving Codex, which is right now the most powerful coding harness. Plus, you can save..."
- 36:36 / Evidence 7: "Make sure to fix this. Read what the codexes are doing. Okay. So what happened is they wrote the same index.html. So now PI agent is going to correct them and you can see that it send the..."
Video-aware target:
- Prompt lane: Coding-agent workflow
- Mechanism to extract: Find the workflow rule that explains when and how to use Codex, Claude Code, browser control, dashboards, or manual review.
- Artifact to produce: A coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal.
- Artifact must include: task class; agent/tool choice; context packet; verification step; handoff/recovery rule
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, transcript support, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable mechanism from the video: Find the workflow rule that explains when and how to use Codex, Claude Code, browser control, dashboards, or manual review. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal.
5. Include:
- a plain-English definition of the core idea
- a diagram or structured model using this sequence: Inspect context -> Route tool -> Plan work -> Edit safely -> Verify behavior -> Report next step
- answers to these source questions: What workflow pain is being solved? | What exact coordination mechanism is shown? | What evidence proves it changes the work?
- 3 concrete examples that apply the video idea to real agentic work, such as rate-limit routing; browser verification after a UI edit; long-running /goal session review
- 2 failure modes the video helps prevent, chosen from the transcript evidence and these likely risks: choosing tools by hype; losing context across agents; letting parallel sessions become invisible
- a checklist for the next real workflow, focused on: routing decision, context portability, verification, handoff summary
- one practical exercise with a clear done signal: Route three recent tasks across Codex, Claude, browser checks, and manual review with a reason for each.
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 "If you don’t run Pi locally you’re falling behind…", not a generic Creative Automation essay.
- Each workflow rule must point to a timestamped claim or demo moment, then state what remains unproven.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- Avoid these generic drifts: generic Codex vs Claude comparison; feature lists without task routing; claims that ignore limits or recovery.
- If evidence is weak or missing, stop and 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.
Creative AI removes the need for taste.
It increases the need for taste because output volume explodes.
The best prompt is enough.
References, critique, iteration, and post-production matter just as much.
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 coding-agent routing and execution matrix with context needed, tool choice, verification, and done signal..
A reusable artifact with a done signal and one verification step.03
Coding-agent workflow teach-back card
Explain the coding-agent workflow mechanism 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.
What distinguishes Pi as a harness from products like Claude Code and Codex?
What are the four ways to improve Pi, from simplest to most powerful?
What is the Pi-plus-Codex workflow and how does it save tokens?
Source shelf
Use the video as a doorway, then verify with primary sources.
