Volo

🧙‍♂️ Building cool stuff with AI -

[Volo @VoloBuilds]

@ericzakariasson Appreciate you sharing a list of actual loops! I've been scratching my head trying to think of loop usecases but I realize now that I was assuming loops had to be infinite. This list with bounded loops is much more practical.

[Volo @VoloBuilds]

What work are AI Agent Loops actually doing? Continuous code review, maybe some research and trend spotting but.. what else?

[Volo @VoloBuilds]

@patrickc I think the coding agents could be decoupled from the real-time collab + context management layer. Then any agent could read/write into it.

[Volo @VoloBuilds]

Like Confluence w/ real-time editing of Google Docs but using plain markdown and possibly has vector search and indexing

[Volo @VoloBuilds]

Any suggestions for a markdown database for agent context? I want a central place for context that both agents can use and non-technical human users can view/modify.

[Volo @VoloBuilds]

@mntruell This is why I'd love to see a duplex voice mode in Cursor. Back and forth conversation with an orchestrator that manages agents under the hood.

[Volo @VoloBuilds]

@NickADobos Reviewing PRs for agent skills and asking people to adjust their phrasing feels so weird lol

[Volo @VoloBuilds]

@MartinTale @shipordie_ I like the aesthetic of the first few but the plain white color of the last one. Here's an example though it might be over-detailed

[Volo @VoloBuilds]

is this really what we've come to?

[Volo @VoloBuilds]

easier said than done of course... More often you'll end up in the weeds on how specific features should work, where to position UI elements, edge cases, etc. Trying to figure out how to overcome this.

[Volo @VoloBuilds]

Information density is a critical but non-obvious lever for building with AI. If you can concisely articulate the essence of your system, AI agents can bootstrap the full spec and self-orchestrate the development.

[Volo @VoloBuilds]

@NuhiAdemaj Did you give it specific skills for the image gen or did it figure it out on its own somehow?

[Volo @VoloBuilds]

Anyone try Antigravity 2.0? How does it compare to Cursor / Codex / Claude Code? Did they fix the 'wipes out your whole system' sandboxing issues?

[Volo @VoloBuilds]

and this is coming from someone who's been using Claude models consistently since Claude 2... Sonnet 3.5 was my absolute favorite for a very long time and Opus 4.5 was another step level change. But Opus 4.7 and 4.8 have been a strange shift..

[Volo @VoloBuilds]

This echoes my experience with Opus 4.7 and now 4.8. It's so adversarial, insisting on challenging basic non-controversial statements and second-guessing everything. Very edgy and kind of paranoid.

[Dylan Field @zoink]

Opus 4.8 is a very strange model. Clearly Anthropic tried to improve honesty, which is commendable. However, the model's curiosity (already worse in 4.7) degraded further. Result is a judgmental personality + sycophancy + sooo much hedging. Basically the opposite of Opus 3.

[Volo @VoloBuilds]

@NickADobos Yeah just gotta have a good tool search so it doesn't preload all of the tool descriptions into context

[Volo @VoloBuilds]

@_MaxBlade What do you have it doing? I haven't figured out any autonomous AI use cases that I'd actually want to be executed continuously

[Volo @VoloBuilds]

@thepatwalls How does someone even find this many customers willing to get on a call? Cold outreach pipeline must have been killer

[Volo @VoloBuilds]

@theo We should have a full-blown RTS game layer on top of agents. Building should feel like playing Starcraft or Civ

[Volo @VoloBuilds]

@omni1896837 @GregKamradt It's actually correct!