This looks really cool for schema migrations but how does it handle updates/inserts if you need to move actual data as part of the migration?

Implementation Notes:

- There is no reason you have to expose the skills through the file system. Just as easy to add tool-call to load a skill. Just put a skill ID in the instruction metadata. Or have a `discover_skills` tool if you want to keep skills out of the instructions all together.

- Another variation is to put a "skills selector" inference in front of your agent invocation. This inference would receive the current inquiry/transcript + the skills metadata and return a list of potentially relevant skills. Same concept as a tool selection, this can save context bandwidth when there are a large number of skills

> Or have a `discover_skills` tool

Yes, treating the "front matter" of skill as "function definition" of tool calls as kind of an equivalence class.

This understanding helped me create an LLM agnostic (also sandboxed) open-skills[1] way before this standardization was proposed.

1. Open-skills: https://github.com/instavm/open-skills

ya... the number of ways to infiltrate a malicious prompt and exfil data is overwhelming almost unlimited. Any tool that can hit a arbitrary url or make a dns request is basic an exfil path.

I recently did a test of a system that was triggering off email and had access to write to google sheets. Easy exfil via `IMPORTDATA`, but there's probably hundreds of ways to do it.

Guys, I think we just rediscovered fascism and social engineering. Lets make the torment nexus on the internet!

Same. For actual productions app I'm typically reviewing the thinking messages and code changes as they happen to ensure it stays on the rails. I heavily use the "revert" to previous state so I can update the prompt with more accurate info that might have come out of the agents trial and error. I find that if I don't do this, the agent makes a mess that often doesn't get cleaned up on its way to the actually solution. Maybe a similar workflow is possible with Claude Code...

Yeah, autonomy has the cost of your mental model getting desynchronized. You either follow along interactively or spend time catching up later.

You can ask Claude to work with you step by step and use /rewind. It only shows the diff though, which, hides some of the problem. Since diffs can seem fine in isolation, but when viewed in context can have obvious issues.

Ya I guess if you have the IDE open and monitor unstaged git, it's a similar workflow. The other cursor feature I use heavily is the ability to add specific lines and ranges of a file to the context. Feels like in the CLI this would just be pasted text and Claude would have to work a lot harder to resolve the source file and range

We have an internal framework for building migrations and the "dry run" it's a core part of the dev cycle. Allows you to test your replication plan and transformations without touching the target. Not to mention, a load that could take >24 hours completes in minutes

My definition of agent has always been an LLM with "effectful" tools, run in a loop where the LLM gets to decide when the task is complete. In other words, an LLM with "agency".

This is exactly how I think of it. An agent has three elements: intelligence (LLM), autonomy (loop) and tools to do anything interesting/useful.

I posted this prediction over a year ago in the Salesforce Reddit and it was an extremely unpopular take[0] (so much so that I don't know if I can post there anymore). Basic argument was that "low code" dsl is far less optimized and accessible for LLM and that there is billions being invested into general purpose code first tooling.

https://www.reddit.com/r/salesforce/comments/1hxxdls/unpopul...

You have no idea what it's like to be American right now. The propaganda information war that's being waged in us is overwhelming and it appears to be working. The world needs to start preparing for a reality where the US can no longer be relied on for security or economic stability. For the sake of all of us, I hope that our European allies are taking serious steps to become more independent from US power and security.

I know there is a lot of good and brave people in the US - I lived there for a long time and call many of your compatriots good friends.

We're trying our best over here, but y'all can't give up at home either. I know it sucks and it's hard, but don't give into the temptation to just tune out. If you don't like what is happening with your country, do your best to change it - don't wait for others to do it for you!

Is 2026 the year of Polish nukes?

I asked a similar question a while back and didn't get any response. Some type of service is needed for applications that want to be AI enabled but not deal with usage based pricing that comes with it. Right now the only option is for the user to provide a token/endpoint from one of the services. This is fine for local apps, but less ideal for we apps.

I do minimal weight training but in climbing the current consensus is that too many reps increases likely hood of developing an overuse injuries in the tendons. Probably depends on the exercises (climbing is hard on the elbows), but maybe keep an eye for tendonitis

Good call out. I’m pretty lazy so I keep the rep ranges low. And not too many sets. Generally I start with a compound lift to hit everything in the muscle group I’m working then move onto accessory lifts to target more granularly. I think I’m lazy enough my risk of injury is low.