SEE CMUX SOLVE Multi-Agent Orchestration (Claude Code and Pi Agent)
IndyDevDan tests CMUX as a solution to three multi-agent orchestration problems — no programmatic access, being unable to monitor to improve, and slow team launch — by giving an agent a CMUX skill and driving a visible fleet of Claude Code, Codex, and Pi agents (running models like Minimax M3 and GLM 5.2) across windows, workspaces, and panes. The theme is scale your compute to scale your impact while keeping every agent fully visible.
IndyDevDan30 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 IndyDevDan; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Skill you build: The ability to orchestrate a visible, programmatically-controlled fleet of heterogeneous coding agents through CMUX so you can monitor, improve, and scale multi-agent work instead of black-box vibe coding.
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.
6,119 cleaned transcript words reviewed across 1,734 timed caption segments.
Thesis
SEE CMUX SOLVE Multi-Agent Orchestration (Claude Code and Pi Agent) teaches a practical coding-agent workflow move: IndyDevDan tests CMUX as a solution to three multi-agent orchestration problems — no programmatic access, being unable to monitor to improve, and slow team launch — by giving an agent a CMUX skill and driving a visible fleet of Claude Code, Codex, and Pi agents (running models like Minimax M3 and GLM 5.2) across windows, workspaces, and panes. The theme is scale your compute to scale your impact while keeping every agent fully visible.
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.
0:50
Three orchestration problems
“fact that these AI labs have massive incentives to keep you and I spending tokens, token maxing, when the truth is there's a dozen different agentic patterns you can use to ship with agents. One of my favorite...”
Problems come before tools. The three multi-agent pain points CMUX is measured against: no programmatic access (you stay the bottleneck, agents can't move at agentic speed), not being able to see agents ('an agent you can't see is an agent you can't improve'), and booting a team by hand killing agentic speed. His preferred structure is three-tier orchestration — an orchestrator prompts leads, leads prompt specialized worker experts — and CMUX gives agentic access to every terminal. Write down your own three orchestration problems and, for each, note whether you currently have programmatic access and visibility into the agents involved.
15:47
Heterogeneous fleets
“who's running patterns we want to replicate, and who's doing stupid we don't want to do again. Okay? And that's on a agent coding tool level, all the way down to, of course, the model level. And then...”
He spins up a security fleet of four coding agents — Claude Code, Codex, and Pi running Minimax M3 and GLM 5.2 — all told to list the top three security vulnerabilities in a repo. The point is scaling compute to scale impact: instead of one agent with sub-agents, run different agentic coding tools with distinct model advantages against the same validation, and unlike a black box, CMUX makes it fully visible so you can see which agent has the edge. Give the same concrete task (e.g. find the top three security issues) to two or three different coding agents at once and compare which model catches what.
19:56
Race to a fix
“This is needle in a hay stack. Capture the flag. first agent to the goalpost wins type of task. Okay. And multi- aent orchestration lets you do this really really well. Every context model prompt, every agent coding...”
For a production emergency, boot an eight-agent race — a needle-in-a-haystack, capture-the-flag, first-agent-to-the-goalpost task — throwing varied compute (Opus, Sonnet, Codex, local Qwen) at the problem in parallel so you take the first correct answer and deploy the hotfix. Notably a couple of Pi agents didn't fire (looked like an env-var setup issue), while the Codex and Sonnet models found the solution quickly — a reminder these new tools still have maturity gaps. Simulate a hotfix by racing several agents at one bug in parallel and take the first correct fix, noting which agents stalled and why.
01
Inspect context
Start with this video's job: IndyDevDan tests CMUX as a solution to three multi-agent orchestration problems — no programmatic access, being unable to monitor to improve, and slow team launch — by giving an agent a CMUX skill and driving a visible fleet of Claude Code, Codex, and Pi agents (running models like Minimax M3 and GLM 5.2) across windows, workspaces, and panes. The theme is scale your compute to scale your impact while keeping every agent fully visible. Treat "Inspect context" as the outcome you are trying to make visible, not a topic label. Anchor it to 0:50, where the video says: “fact that these AI labs have massive incentives to keep you and I spending tokens, token maxing, when the truth is there's a dozen different agentic patterns you can use to ship with agents. One of my favorite...”
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 15:47, where the video says: “who's running patterns we want to replicate, and who's doing stupid we don't want to do again. Okay? And that's on a agent coding tool level, all the way down to, of course, the model level. And then...”
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: IndyDevDan tests CMUX as a solution to three multi-agent orchestration problems — no programmatic access, being unable to monitor to improve, and slow team launch — by giving an agent a CMUX skill and driving a visible fleet of Claude Code, Codex, and Pi agents (running models like Minimax M3 and GLM 5.2) across windows, workspaces, and panes. The theme is scale your compute to scale your impact while keeping every agent fully visible.
02
Explain the practical stakes without hype: New playlist item from IndyDevDan; 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: SEE CMUX SOLVE Multi-Agent Orchestration (Claude Code and Pi Agent)
- URL: https://www.youtube.com/watch?v=WAFUMBLOjHo
- Topic: Creative Automation
- My current learning frame: In CMUX (or tmux), give an orchestrator agent a CMUX skill and have it launch a small heterogeneous fleet — e.g. Claude Code, Codex, and Pi — to race the same security or bug-finding task, then watch each agent to see which behaviors to reinforce or cut.
- Why this matters: New playlist item from IndyDevDan; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Transcript anchors from this exact video:
- 0:50 / Evidence 1: "fact that these AI labs have massive incentives to keep you and I spending tokens, token maxing, when the truth is there's a dozen different agentic patterns you can use to ship with agents. One of my favorite..."
- 3:39 / Evidence 2: "between vibe coding and agentic engineering. An agent you can't see is an agent you can't improve. I need to be able to see everyone of my agents. Okay? It doesn't matter if it's cloud code, pi, open..."
- 7:02 / Evidence 3: "place to finish, right? We can do a lot better than that with the right agentic tools. That's why we're trying to see if we can really improve our multi- aent orchestration abilities. Okay, so that's the mental..."
- 14:04 / Evidence 4: "new fleet. This is our security fleet. Send up four coding agents. Cloud Code, Codeex, Pi with Minamax and GLM. Create a fleet. List the top three security vulnerabilities you can find in this repository. So this is..."
- 15:47 / Evidence 5: "who's running patterns we want to replicate, and who's doing stupid we don't want to do again. Okay? And that's on a agent coding tool level, all the way down to, of course, the model level. And then..."
- 19:56 / Evidence 6: "This is needle in a hay stack. Capture the flag. first agent to the goalpost wins type of task. Okay. And multi- aent orchestration lets you do this really really well. Every context model prompt, every agent coding..."
- 27:23 / Evidence 7: "monitoring agents so you can improve them. And then it's being able to scale your orchestration to whatever level you need to. And so, as mentioned, this is not top-down agent communication, right? Any agent can prompt any..."
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 "SEE CMUX SOLVE Multi-Agent Orchestration (Claude Code and Pi Agent)", 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 are the three multi-agent orchestration problems CMUX is measured against?
What is the security fleet, and what is the point of running different coding tools on the same task?
How does the production-emergency pattern work, and which agents stalled?
Source shelf
Use the video as a doorway, then verify with primary sources.
