Wait. Cursor had access to the production DB???

I mentally replace “AI agent” with “intern” when I read this stuff and it helps clarify the root cause. People are connecting directly into prod and making changes live? It’s not (solely) the fault of the actor, but of the whole process that makes it possible for the event to happen. It may be the case that there’s a break-glass situation where a specific person needs prod access to fix an urgent thing. In that case, there needs to be an approved plan like “I’m connecting to this DB, making this query to find the affected row, then running this one to fix it”.

If it makes you shudder to imagine allowing an intern to do a thing, you should shudder harder to imagine letting an AI — an intern who can type really fast — do it.

I work in AI. I love using AI. I don’t want to go back to not using AI. But darned if I’m letting anyone, human or AI, just waltz into a prod environment and make random changes.

Does the computer running the agent have production DB credentials on it anywhere? If it does, the AI has access to the production DB.

This is part of why I'm bearish on the new hotness of "don't write tools, just write a Markdown skill and let the LLM write its own bash commands". It does work, for the most part, at the cost of it being entirely capable of changing its environment and executing arbitrary commands. Approvals exist, sure, but I've never seen anyone manually approve a command past like the 3rd permission dialog.

It didn't have access to any db. In short: It went looking in the codebase for a credential to manage the staging environment, found a testing credential unrelated to anything it was doing, that the devs didn't know had permissions to administer anything, and then used that to delete the wrong db.

I would never eat burnt toast, but dark roast coffee, a pint of Guinness I wouldn't think twice. How bad are these comparatively?

There was a problem providing the content you requested

Bad bot

I have the same issue with a Boox eink tablet. I`m pretty sure I`m not a bot.

But how can you be sure, really?

So if I release a new cli. How do I get the LLM to know about it? Do i tell it every time to run the command? Do I build a skill. Should I release a skill with the cli? Do I just create docs on GitHub and hope the next crawl gets into the training set?

Package a skill with your CLI itself and give users instructions on how to install the skill properly. That allows the agent to read the instructions in a context efficient way when it wants to use the CLI

Just have to ask. Will I be spending way more money since my context window is getting so much bigger?

Yes, full context is used to generate each new token.

Doesn't matter what we want or how we feel. Product, C-suite, Customers just want software as fast and as cheap as possible. They don't care about the code and the craft. If that's the case then we have to use AI if we want to stay marketable.

I wonder if customers even appreciate the organic artisanal labels that some sites are putting up e.g. https://play.date/games/diora/

I just paid for it, downloaded the kindle app, and dragged the .awz8 file and it says it doesn't support the format!

I think you might be doing something wrong then? Maybe you have the wrong file? The Kindle For PC app should always give you azw files, not azw8 (even if it does use the newer format they're all just named azw).

It should show up in the epubor app on the Kindle tab after you install the kindle app and used it to download your books. No need to drag and drop from the file system it's all right there in the app. It finds Kindle, Kobo, etc and lists them.

EDIT - Make sure you leave the Kindle app running. I think it needs to be able to read the keys from memory or something.

How do I get to chapter 2!

Coming soon! I about a month or so

Live github PRs that you are tagged in.