Google just released a free 2-hour+ course on full agent engineering. How to go from one prompt to a system that runs while you sleep: 38:46 - Build your first AI agent 54:46 - Connect MCP tools 1:12:43 - Run agents with four loop patterns 1:20:57 - Turn agent... loops into graphs 2:22:31 - Build the complete autonomous system Most people are still building one agent and stopping there. Google is already teaching the full stack: Agents → Tools → Loops → Graphs → Autonomous Systems Single agents are the old workflow. Systems that run without you are the new one. This course is worth more than most paid agent engineering bootcamps. Bookmark and watch it today Then read the full graph engineering playbook belowshow more

Moving from single agents to autonomous multi-loop systems is the ultimate meta-upgrade

Agent engineering roadmap Seriously valuable drop

If you're building agents from this course and want to test how different models handle the same tool-calling loops, has unified API access to GLM 5.2, DeepSeek V4 Pro/Flash, and Kimi K2.6 with keys and channel management ready to go.

That's awesome! Definitely going to check this out. Thanks for sharing!

The timestamp order is what I like here, tools first, then loops, then graphs, it makes the architecture feel earned instead of starting with a diagram and hoping the pieces fit

AI development is moving from building single agents to designing full systems with tools, memory, and feedback loops. I read about it in @TheUpsideAI this morning.

i like courses by Google, they're very informative

@grok when was this video linked in the repost originally published by google? The article looks like it is slip streaming on a repost of tangentially related content.

四种 loop pattern 这段最值得先看

ngl the agent loops part is probably where most people get stuck. curious if they actually show real examples or just theory

free high quality courses from google are always appreciated

Pretty nice

Wow, this is amazing! Thanks for sharing this, I've been looking for something like this.

The loop to graph transition is the real unlock here. Most people stop at basic chains, but that workflow structure is what makes agents actually reliable while you sleep. Good move from Google dropping this for free.

@Josefusan111

Nice

this is more than a post, it's something people are going to keep and use — that's what good content looks like

The useful jump is from “agent loop” to operating model: what can it touch, what evidence must it leave, who reviews failures, and when does the loop stop. That’s where demos become production systems.

Appreciate that they included timestamps. Makes it much easier to jump into the parts you actually need. Cool info sanks

Locked in

One of the most overlooked aspects of climate change modeling is the trade-off between efficiency, security, and cost under resource constraints.

@grok youtube link

'runs while you sleep' always sells the course. mine runs while i sleep too, i just wake up to whatever it confidently broke at 3am.

Great resource. The gap between 'agent that runs' and 'agent that runs reliably in production' is still massive — access control and guardrails make all the difference.

this info shouldn't be for free

2+ hrs covering the whole stack is a lot of scope

And this is here in portuguese:

love Google courses, and I’m definitely not skipping this one