Aaron Diamond-Reivich

@j_becke Looks awesome! I love that I can change the theme, fun to play around with

New website just dropped (please help us find rough edges!) macro.com

Being able to easily share these flows with the rest of your team is a huge deal. When we onboard new teams to Mito, usually one person takes on the role of context creator. They: 1. write up custom rules like: If you want churn data, merge these tables on these columns, etc. 2. try to teach the rest of their team how to use the custom rules, how to keep them in sync, etc. For teams that already use Notion, some of that context, and all of that collaboration piece is already taken care. Huge ops win.

Mainly: Notion custom agents can be shared with a team. Worker tools can be versioned in git with auto-deployment through Actions. Custom agents have a run history that can be audited by other people to understand what the agent is doing across complex systems and tooling.

@arvinds @abacusai Testing now

@_aaronDR @abacusai Should be fixed with our push earlier today

.@abacusai, @arvinds can you confirm how to set the response format for the LLM router: - /app/route-llm-apis says to use {"type": "json"} - /help/developer-platform/route-llm/api says to use {"type": "json_object"}

@arvinds @abacusai Appreciate it. Do you know if that response type is passed through to non-openai models. I opened an issue here github.com/abacusai/api-p…

@_aaronDR @abacusai Thanks for noticing - we are fixing /app/route-llm-apis - the developer-plaform {"type": "json_object"} is the correct way

Anyone else noticing that gpt-5.2-codex is prone to returning the complete response_format object twice in one completion? Haven't seen this issue with other models from openai

My new favorite hobby is trying to trick Apple AI into generating text summaries that have nothing to do with what I actually said

@leerob Version 2.4.31. I tried using the integrated browser with @browser and just telling it to use the Chrome tab that I have open as well

@_aaronDR What version are you on? And is this using the integrated browser inside of Cursor (e.g. @​browser)?

Getting Cursor to verify its own changes is such an unlocked workflow, but recently I've been having some issues getting Cursor to use the browser. Any ideas on what I'm doing wrong here @leerob? Appreciate any help greatly!

@thdxr "You get to look at and take pride in the code you write" is going to be a new benefit that companies can list right next to unlimited PTO

a lot of our recent hires care deeply about code how's it's written, how to make it elegant, how to make it easy to understand and a joy to maintain it's a counter bet in a field where most people are saying we won't be looking at code for much longer

@marclou @_MaxBlade Would be curious to see a graph of MRR vs X followers. Both at the time of launch of the product and over time.

My new hobby is watching my friends build profitable startups in real time. ⁦0 to $4K MRR in 36 hours for @_MaxBlade⁩ 🤗

Is @LiteLLM Cloud down? Can't access my admin portal

When I ask Cursor to update its plan with Todo's, it often create a new #TODO section in the markdown file rather than filling out the todos yaml that populates the launchable todo list at the bottom of the plan. Any tips on how to prompt this correctly?

@ryannystrom I wish Cursor had this too. Sometimes it's very hard to find my old chats to reference. Would be nice to pin chats or plans that I know I'll want to come back to later.

The little details matter

Looks like 5.2 isn't going down without a fight

Asking gpt-5.2-codex to add support for 5.3-codex in your product feels like asking the person you're about to fire to train its replacement

@simpsoka @bentlegen Sort of ironic that this is coming from the same company that has a 'request silent ride' option in their app

@bentlegen I saw a sign on a uber recently that said, “real driver,” like it was a good thing.

Last week I took an Uber in SF and the driver thanked me for not taking Waymo Said he used to earn $400-500 on a Friday night - now he’s lucky to hit $300

@arakharazian @Lovable @tryprofound I'm always surprised when Ramp, OpenAI, etc share info about their most successful customers like this. Of course it's good publicity, but still feels strange. Would be curious if things like this get sent to the companies for approval first.

Fastest growing companies on Ramp trends to watch: (1) AI infrastructure means open source is likely ticking up, (2) AI dev tools (@Lovable) are not going away, (3) AEO is the future of marketing (@tryprofound)

Initial impression is Opus 4.6 is way better at making changes, then using the browser to verify its work and iterate than Opus 4.5 With 4.5 it was always a fight to get it to use the browser. 4.6 has been iterating without prompting on an FAQ design for the past 10 mins.