Easy if you ignore the security aspects. You want to hand over your tokens to your LLM so it can script up a tool that can access it? The value I see in MCP is that you can give an LLM access to services via socket without giving it access to the tokens/credentials required to access said service. It provides at least one level of security that way.

The point of the example seemed to be connecting easily to a scoped GraphQL API.

It won't matter. The core ideas of an Agent Manager view will be copied and improved by others in many project in the future.

I haven't used Windsurf (been using Claude Code and similar). Does it provide an Agent Manager window/view? This to me looks more useful to me than the browser integration piece.

The Agent Manager view providing a unified view of all active agents and allowing you to immediately respond to any approval requests or followup questions looks very useful regardless of which VCS you're using under the covers. Am I missing something here that jj does?

See my setup detailed in another sibling comment of yours, jj is just a small part of it, and you can probably get that with git too.

I’m already at full mental capacity planning and reviewing the work of two agents (one foreground which almost never asks for approval, and one background which never asks for approval).

I don’t really need the ability to juggle more of them, and noticing their messages is not a bottleneck for me, while I’m happy with the customizability and adaptability of my raw’er workflow.

Maybe if they’re as slow as codex…

Fair enough. I use git worktrees (with a script that creates the git branch, worktree and opens a new vs code workspace). You're right, managing more than about two active sessions at once is probably the limit though I'm somewhat hopeful that better tooling similar to the Agent Manager window here would allow me to scale a bit past that especially if some of those sessions are more design explorations.

The Claude Code extension on VS Code does very little (too little in my opinion). The integration level with agentic functionality provided by Antigravity goes much deeper in my 20 minutes or so of playing with it. The biggest value pieces I see is: Agent Manager window which provides a unified view of all my agents running across all my workspaces (!) where I can quickly approve or respond to followup questions and quickly brings me to the code in context for each agent, additionally, I can select a piece of code and comment on it inline and that comment gets sent to the correct, active agent. These two things alone are items which I have been looking for... Too bad I only have approval to use Claude Code at work. This looks promising.

That’s all very silly and unnecessary in my opinion. Agentic change by change diffs are optimal for a professional.

Well, you are entitled to your opinion but many people would disagree with you, and that's the crux of the issue, everyone has their own conflicting views on what the UX should be, hence all the forks.

My experience has been the opposite. Oracle (the database) is actually a really solid product for the most part. Oracle (the company) is a different story. My eyes were really opened to some of the technical shortcomings in Postgres when we migrated from Oracle to Postgres a few years ago at $DAYJOB. Things like: a) global temp tables (there's an open source extension we had to use to fake this out), b) RLS (exists in PG but most functions that you might need to build on top perform badly), c) crashes in PG take out the whole database and a host of other smaller items. I'm not saying it wasn't worth it, but I wouldn't pretend Postgres is the best database either.

Oracle the database is the Apple of Enterprise. They have some best in class aspects, but also a lot of you're holding it wrong. They cost a lot and have a really arrogant culture. They both have a rabid reality-denying fanbase. The support is either fantastic if you ask a known question, or a hellscape of inconsistent answers and denials if you dont.

They lost me decades ago over the lack of auto incrementing PKs and having to cobble them together through a sequence and a trigger if I remember right. Seemed like utter nonsense. But I'll take your word that you got value out of the features you mentioned. The company side will forever prevent me from taking any of their products seriously though.

I love Postgres and it is a shining example of how good software can be. Great job everyone, no notes.

BUT Oracle has some killer features that PG just doesn’t. The first that comes to mind is for-real multi-master. Close second is declaring partitions in the DDL of the table itself.

You would typically want to use the same database instance for your queue as long as you can get away with it because then transaction handling is trivial. As soon as you move the queue somewhere else you need to carefully think about how you'll deal with transactionality.

I believe most setups using DB+{Queue,Kafka} don't truly deal with it fwiw.

Also add `"includeCoAuthoredBy": false` to your `settings.json` file (you may also need to reinforce this in your commit prompt YMMV).

ahhhhh thank you! this saves me from having to add this to every repo's CLAUDE.md file.

We definitely have this system in place in some cities in Canada, primarily for express bus routes.

So as a driver, you want to follow an express route bus when you can?

I tried Happy Coder for a bit. It seemed exactly what I was missing but about 1/2 the time session notifications weren't coming through and the developers of the tool seem busy pushing it off in other directions rather than in making the core functionality bullet-proof so I gave up on it. Unfortunate. Hopefully something else pops up or Anthropic bakes it into their own tooling.