OpenCode: The #1 Coding Agent, 8 Million Devs Chose Over Claude Code

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

Source video:
- Title: OpenCode: The #1 Coding Agent, 8 Million Devs Chose Over Claude Code
- URL: https://www.youtube.com/watch?v=KvFmNSP7nFw
- Topic: Interfaces + Open Design
- My current learning frame: Install OpenCode, connect a provider and run /init to generate agents.md, define a custom sub-agent on a cheap model for one slice of work, then compare its output and cost against your usual flagship-model coding tool.
- Why this matters: New playlist item from AI Stack Engineer; queued for transcript-backed review, topic mapping, and a practical learning artifact.

Transcript anchors from this exact video:
- 0:34 / Evidence 1: "walk through the entire Open Code ecosystem. Show you how the pieces fit together and figure out where it actually beats tools like Claude Code and Cursor and where it still loses. Quick bit of context first because..."
- 2:14 / Evidence 2: "Once it's installed, you cd into your project, type open code, and you're in the terminal UI. From there, you run the /connect command to hook up a provider. You paste an API key, and that's it. Then..."
- 3:48 / Evidence 3: "its own permissions, and its own system prompt. So, for example, I have a front-end agent running GLM 5.2 because that model performs really well on design benchmarks and a code review agent running a clawed model with..."
- 5:26 / Evidence 4: "screen, and it even connects to remote machines, so the agent runs on your server while you drive it from your laptop. On Windows, it talks to WSL under the hood. The point is, you no longer need..."
- 7:04 / Evidence 5: "outran capacity. So, a free open-source project quietly built a revenue model that competes directly with Cursor and Copilot subscriptions. And it did that without locking anyone in because Zen and Go Keys work with other agents, too."
- 8:48 / Evidence 6: "model gateway, and paid plans. While cursor got acquired, Windsurf got split up and Gemini CLI got shut down. The independent open- source option turned out to be the stable one. That was not the outcome anyone predicted,..."

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 UI critique sheet for judging whether an AI interface improves control.
5. Include:
   - a plain-English definition of the core idea
   - a diagram or structured model using this sequence: Intent -> Canvas -> Artifact -> Preview -> Feedback -> Iteration
   - 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 "OpenCode: The #1 Coding Agent, 8 Million Devs Chose Over Claude Code", not a generic Interfaces + Open Design 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.