OpenAI DevDay 2026: a cheaper Sol, Dots and Codex with your laptop closed
Close your laptop and let Codex keep working. Add a cheaper Sol and Dots for recurring tasks, and there is plenty to unpack from DevDay. Here is where we would start.
News and guides about AI agents, n8n, MCP and evaluating results. Plan workflows with room for human decisions.
Choose a draft or an information-sorting task first. Plan where approval belongs before connecting a live account.
Close your laptop and let Codex keep working. Add a cheaper Sol and Dots for recurring tasks, and there is plenty to unpack from DevDay. Here is where we would start.
The assistant quoted a ticket price. The trouble is, the source had no price list. We have twelve such situations: see how JEV, GPT-6 Luna and you judge them.
One model answers; another judges it. Sounds sensible, until we ask who checks the judge. We take a look at JEV and share an exercise of our own.
The assistant knows the conversation, but the order status lives in another system. MCP helps connect the two. Let’s look at an ordinary question about fulfillment.
The document was supposed to provide information. Instead, it tries to direct the assistant. We explain prompt injection with a simple example and look at where to set boundaries.
Running a model on your own hardware sounds appealing. With Gemma 4, we start with one task, then look at variants, memory and everyday maintenance.
AI writes the reply; a person decides whether to send it. Let’s build that process in n8n, with the full text to review and a plan for rejection, silence and retries.
Someone gathers sources, someone writes from them, someone revisits the suspicious sentence. Let’s divide the work into those three roles, even if one person takes all of them.
You message an assistant in a chat app, and it uses connected tools. That was the idea behind Clawdbot, now developed as OpenClaw. We explain it and correct a common shortcut about local data.
Instead of “press one,” a conversation about an available appointment. We look at what voice agents need to book a visit and how to tell whether they actually help.
A folder full of material and a report to prepare. Cowork lets you delegate file work in ordinary language. Here is how to define the task and check what comes back.