I Was The Only Thing Connecting Claude, ChatGPT, and Codex. So I Built My Replacement.

You are helping me turn one specific YouTube video into real, durable learning.

Source video:
- Title: I Was The Only Thing Connecting Claude, ChatGPT, and Codex. So I Built My Replacement.
- URL: https://www.youtube.com/watch?v=QSK4vf_ZTRA
- Topic: Creative Automation
- My current learning frame: Stand up a free Linear queue, install a simple agent-instructions protocol, and run the smoke test — create a 'say hello from the queue' issue, have one agent claim it, move it through agent-working to done with a receipt — then route one real cross-tool task through it.
- Why this matters: New playlist item from AI News & Strategy Daily | Nate B Jones; queued for transcript-backed review, topic mapping, and a practical learning artifact.

Transcript anchors from this exact video:
- 1:09 / Evidence 1: "a baby. She runs an agency. She uses Claude code. She's got loops and automations. She's looked seriously at OpenClaw because she wants agents that do real work. She is not trying AI for the first time. She's..."
- 6:32 / Evidence 2: "as easy to use and do as possible. Just put the work in a queue that both people and agents can read. What is a queue? It can be as simple as a Jira system. It can be..."
- 9:55 / Evidence 3: "you're not limited by only using those tools. Open Claw is pointed at a real desire that made a lot of sense. We want agents that can act, right? Agents that aren't a chat window. Hermes and similar..."
- 13:18 / Evidence 4: "agent claimed receipt. And then execution starts locally, right? Linear coordinates the team. Codex does the work. A human can create a task for that agent and an agent create a task for another agent. The task includes..."
- 15:49 / Evidence 5: "resume and the audit trail stays in one place. And when finished, Leo leaves agent done as a status and moves on to the next task. And we can use that for any piece of work. You can..."
- 17:24 / Evidence 6: "draft the two messages that might be needed to the school and wait for approval before anything leaves the system. The model can be the same here, but the assignment is much more clear and the ability to..."
- 20:16 / Evidence 7: "clean framework so agents can carry work to the point where judgment is needed and not bother us for the annoying handoffs along the way. That's the version of autonomy I actually want. That's what I've been living..."

Video-aware target:
- Prompt lane: Hermes operations
- Mechanism to extract: Identify the operations control that makes long-running agent work visible, recoverable, or safer.
- Artifact to produce: A Hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria.
- Artifact must include: health check; state model; permission boundary; log source; recovery action

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: Identify the operations control that makes long-running agent work visible, recoverable, or safer. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A Hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria.
5. Include:
   - a plain-English definition of the core idea
   - a diagram or structured model using this sequence: Project state -> Session -> Queue/Kanban -> Tools -> Logs -> Recovery -> Post-run review
   - answers to these source questions: What operational failure is prevented? | What state is visible? | What can be recovered or redirected?
   - 3 concrete examples that apply the video idea to real agentic work, such as Hermes Kanban triage; local model endpoint check; agent swarm recovery review
   - 2 failure modes the video helps prevent, chosen from the transcript evidence and these likely risks: treating UI features as reliability; missing logs; no stop/recover path
   - a checklist for the next real workflow, focused on: status, model/backend, tools, logs, recovery
   - one practical exercise with a clear done signal: Write a runbook for restarting one stuck Hermes-style agent session.
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 "I Was The Only Thing Connecting Claude, ChatGPT, and Codex. So I Built My Replacement.", not a generic Creative Automation essay.
- Ground each ops recommendation in transcript evidence about state, queues, models, tools, security, logs, or recovery.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- Avoid these generic drifts: feature cheerleading; ops advice without logs/state; assuming reliability from a demo alone.
- 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.