the reverse brain drain is complete
5% is a bit surprisingly but honestly makes sense. People massively underestimate the long tail of free users, abuse, and fraud in coding harnesses
The $20/mo tech bro on X is not representative of actual inference costs. Power users can burn multiples of their subscription, and there are elaborate schemes built around arbitraging subsidized model access
Cursor’s economics are increasingly centered on its own Grok/Composer pool, with third-party models metered probably closer to API cost
We’re sorry to see that OpenAI put out a note saying they plan to block Cursor users from accessing OpenAI models in three months.
OpenAI models serve about 5% of Cursor user traffic, and we’re speaking with the OpenAI team to resolve this.
Cursor was one of the very first users of OpenAI, we’ve worked closely with their team for years, and we’ve trusted their platform to be neutral infrastructure for our business.
Hong Yi Chen retweeted
For the past two days @warpdotdev has been the number one trending repo on @github. Now almost at 50k stars, double where we were 48 hours ago. Really psyched to be building with the community + agents.
Thanks to everyone who is hopping in and contributing. We have landed a ton of community PRs already.
Hong Yi Chen retweeted
You can now run any CLI agent with first-class support in Warp, including Claude Code, Codex, OpenCode and Gemini CLI.
• Vertical tabs
• Notifications when they need you
• Integrated code review
• Remote control from mobile
• Rich input editor
Download Warp for free today.
Hong Yi Chen retweeted
We moved our docs site from a hosted solution to Markdown with @astrodotbuild in a matter of hours, made them better, and saved $10k+/year.
@BHolmesDev, @petradonka and @hongyichen led the migration:
- Ported over all 266 pages, including video embeds, tab components, callouts, and image formatting
- Built a custom Ask AI widget backed by Kapa that also improves the experience
- Exposed all docs content as plain Markdown for agent and LLM accessibility
- Set up hosting on @vercel, including proper preview deployments and better analytics
Now, they're using computer use to validate that all pages migrated without regressions. Should be in prod soon.
ALT Merged commit shows migration of port terminal documentation, featuring 873 files changed with significant additions and a few deletions.
Hong Yi Chen retweeted
@amrcn_werewolf is VP of Core AI at Microsoft and has figured out how to make the terminal his command center for absolutely everything.
From assigning Azure IAM roles to reconfiguring his daughter's homework, @warpdotdev + the terminal has become his "ad hoc agent army."
In this ep he show us how he:
- Uses the Azure CLI to assign complex roles in seconds instead of an hour in the portal
- Activates his scanner remotely--no button pressed
- auto-schedules meetings the moment someone requests a time with @Copilot
This one is jam packed with tips, hacks, and shortcuts to make work + life easier.
As always--tysm to our amazing sponsors:
🐶 @Atlassian Rovo - AI that knows your business: rovo.com/
💜@Lovable - build apps by simply chatting with AI: lovable.dev/
Watch the full episode: youtube.com/watch?v=diy3kmUl…
two takeaways from judging @ treehacks this weekend
(1) when building is a commodity, bad ideas get exposed fast. the signal vs slop gap becomes obvious, and it’s now rare to see a devpost submission that isn’t packed with AI-written copy
(2) AI makes demos cheap, so hackathons shift from demos → outcomes. Prediction: hackathons + incubators + case comps blur into a make-a-thon: ship, talk to users, iterate, and leave with traction (or proof it’s dead)
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
just one more agent bro
it’ll refactor the monolith bro
for real this time bro please bro
i promise it won’t touch prod bro
please bro just one more agent bro
trust me bro it’s just a quick cleanup
We built Oz because local agents hit limits fast: CPU/memory, repo sprawl, and no shared visibility
Oz turns agents into infra: consistent environments + orchestration + API/SDK, so teams can build real systems (triage, incident response, PR workflows), not just interactive chat
Introducing Oz: the platform to orchestrate agents in the cloud.
Spin up hundreds of agents from your terminal, browser, the API, or your phone. Each agent gets a @docker environment to build, test, and write PRs.
Come back from your lunch break to code that’s ready to merge.
