vist dev

Secure founder pricing now, before it runs out → up with my thoughts →

[Thariq @trq212]

the right way to use model capabilities is not to ship 10x more features to prod

it's to spend more time understanding your users, trying experiments, building prototypes, learning about things you don't understand so that you can ship things that actually work

[vist dev @vistdev]

@danellisona @Zallordian Not always 🫣… sometimes the conversation turns out to be with a marketing bot. 😞

[vist dev @vistdev]

@philaraujoPM Great observations! The solo company-brain attempt usually dies on maintenance (like most things, right?) ... so until the AI can read/write the notes without you babysitting every paste, it's doomed!
Are you making any difference between notes and decisions and so on?

[vist dev @vistdev]

@vinsonleow @grok @claudeai @AnthropicAI @ChatGPT @OpenAI Makes you wonder what they are _not_ shipping, but have been cooking internally...

[vist dev @vistdev]

@vinsonleow @claudeai @grok @ChatGPT @GeminiApp @Kimi_Moonshot Or: use Vist, with zero setup beyond a simple remote MCP connector, and get it all out of the box, no extra setup or tokens needed ;-)

[vist dev @vistdev]

This I am also very much looking forward to. Opus is still my go to for making plans…

[Claude @claudeai]

Opus 5.5 requires less compute to serve than Opus 5, and its pricing reflects that.

Our tests show that at default settings it will cost 40% less than Opus 5 on typical workloads.

[vist dev @vistdev]

I do hope the first reactions hold up and it’s more concise and down to earth than 5…

[Claude @claudeai]

Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.

It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.

[vist dev @vistdev]

You open a fresh chat and re-explain your project, what not to do, and what you were mid-way through on Friday.

If that sounds familiar - it's not a model problem. You need a shared brain for you & the AI.

MCP, markdown, EU-hosted, simple yet efficient.
https://t.co/Ou3UcJ7hII

[claire vo 🖤 @clairevo]

so i have 40+ agents who go through my various emails, crms, slacks, etc and finds all the things i need to do.
they tell me to do it, and i ignore it.
so i vibe coded an app that gives all the agents a central place to tell me all the things to do.
i look at it, and i ignore it.
so then i open the app in codex and i tell codex to bully me into doing it...

[vist dev @vistdev]

@thedavidqu For me, I found that because I like to drive work from my phone as well, I needed something that is at the same time simpler (less setup, hosted) and more advanced (memory features) than Obsidian.

[vist dev @vistdev]

@kodevbuilds that really is the end goal, yes... one thing that I can run from any LLM I choose to work with, and gives the LLM all the knowledge and tools to help me run all my work

[vist dev @vistdev]

@mignano @ujjalcal That’s my feeling as well, and I’ll need to experiment with it asap. The pieces are there for us to combine them in new ways.

[vist dev @vistdev]

@vernons @volsokotun Absolutely, and also: benchmarks don't really matter, no one needs the highest score. It just needs to meet your requirements, no more...

[vist dev @vistdev]

@vernons My healing process: AGENTS.md has static rules, but what ended up making the difference for me was using a Markdown notes/tasks app that both of us (AI via MCP, me in its minimalist UI) write to, so the next harness auto-loads context. Smoooooth... ;-)

[vist dev @vistdev]

@mateusz_palak Many use cases that people talk about - saving bookmarks? youtube video urls? - seem bit pointless for sure. But as a product person in a large org, having my agents organize, write, and pull together information regardless of whether I am on my phone or desktop... huge win!

[vist dev @vistdev]

@thdxr @X So true... it's turned me off Reddit already and brought me back to X, but if it gets any worse here... what's left? IRC?

[vist dev @vistdev]

@AlexFinn But how do we get the enterprise to follow suit? Try getting a Linux laptop as your main workstation at a typical corporation... running into a concrete wall of IT says no

[Peter Steinberger 🦞 @steipete]

New benchmark dropped https://t.co/5lFZNcDNCh

[vist dev @vistdev]

@BowTiedMarmott @Trey_Harnden This is exactly the kind of scenario I built Vist for. All my AI agents (Claude/Cursor/ChatGPT/...) on all my devices (web/desktop/mobile/terminal) connect to Vist via MCP and share the same memories. And because Vist is minimalist markdown notes/task app, I can read it all.

[vist dev @vistdev]

@aiwithremy How are you instructing your agents what to use when? I find a simple line at the top of my agents.md to use specific tools for specific purposes does the trick reliably.