Yes this was my question as well, an example showing all the plumbing/dependencies to generate a graph from an external database server would be very helpful.
We certainly plan to create a few videos showing how to set it up and use it. If you use it in Positron with the ggsql extension it can interact directly with the connection pane to connect to the various backends you have there
Cloudflare is spending years of goodwill earned through technical skill, trending towards AI enshittification starting with their blog posts and vibe coded features/products.
I also kind of rolled my eyes at the blog post and its obsessive focus on "agents" -- definitely feels like a solution looking for a problem. But the email-sending product being promoted is probably ok, right? They just happened to write a lot of words observing that ChatGPT can, in fact, call sendmail() through their platform (if you give it access) -- a fact that shouldn't surprise anyone.
Our initial blog covered most of Email Service's API and what you can expect from it in terms of deliverability, DNS records setup, etc. https://blog.cloudflare.com/email-service/
Email Service can definitely be used as a transactional email API, and it has everything you would expect like SDKs, binding, observability and more coming on the way
The agent angle in this post reflects what we're actually seeing from developers during our private beta. And the idea that an agent can have an inbox to communicate is a new piece in the developer toolbelt.
Thanks for the clarification! Sounds like some developers, including your beta users, are experimenting with new ideas (which includes plugging agents into different workflows to see what happens), while old farts like myself bemoan AI getting plugged into everything and every app sprouting "Ask AI" buttons that they never asked for or wanted.
I can definitely understand some of the ire-- people are probably imagining how they'll try to contact Verizon and will get back a totally unhelpful email from ChatGPT when all they wanted was to talk to a real human for 5 minutes. Your blog post about hooking up agents to email probably speaks to that fear.
It's most likely people doing active development, willing to experiment on a test/beta platform... which probably correlates strongly with those testing/trying agent based workflows.
Thomas can probably speak to this better, but as someone who has participated in other Cloudflare betas: there's usually a button or a form and you can request access.
It's like the author handed the copy to the editor who then added a new broken sentence after each original sentence that somehow jams "agents" in there.
R.I.P. to the Amazon Glow video calling device, killed before AI went mainstream. I'd love to hear how to get root on one... exactly the hardware your project could use most effectively and an amazing interface for playing games remotely with the grandparents.
What you may be remembering were reports of exceptional cases where it didn’t handle out of memory errors well. I was one of the people affected. I was running complex analytic queries on 400 GB parquets and I only had 128GB memory. It used jemalloc which didn’t gracefully degrade. They fixed a lot of the OOM issues so it’s more robust now. I haven’t had a crash for a long time.
PSA: This is the best place to collect upvotes for your vibe coded ideas/projects that you think might not be up to "Show HN" quality yet, whether reproducible at the source code or prompting level(s) of software development or not... the bar is understood to be much lower here.
Surprised to see no mention in the article or discussion yet about using MCPs in 'code mode', where an API is generated client-side relying on MCP primarily as an interface standard. I'm still learning but I've read this reduces the amount of context required to use the MCP.
It seems like a lot of the discussion is arguing in favor of API usage without realizing that MCP basically standardizes a universal API, thus enabling code mode.
reply