The Product-Led Geek | Founder: https://xitter.cf/t.co/R8QTNLzYG0 | https://xitter.cf/t.co/kqy9odnI7Z | Dev tool growth advisor: https://xitter.cf/t.co/MGroaTRN7A

Joined June 2010
This is proving to be such a nice way to stay on top of what all the clankers are doing.
2
16
2,534
TBC on real world but these jumps are crazy. Chapeau @Kimi_Moonshot
169
This is making working with lots of agents much easier. Coming soon.
1
281
I’m all for it. 🔥
alright, it's a good time to finally just say what we are doing: we are rebuilding the parts of github that no longer work. depot ci was one step in that direction, switch to it now if you're impacted by this latest outage. depot code is what comes next: our take on source control hosting, with our take on how to build trust around the increased velocity of code being written. i'll be the first to say that this isn't a novel idea. we are not alone in seeing the pain points we are all facing. many folks are seeing the same thing, hell depot ci integrates with an increasing number of them. that's not an accident. the truth is that nobody quite knows what this all looks like a year from now, let alone three years from now. but we are all collectively tired of the status quo and can very clearly see that the way we work with agents is fundamentally different. and thus needs a different approach than what we've used to date. furthermore, it's fucking exciting to see that there is even an opportunity to rethink how this is all done. i think that is refreshing and i'm excited to see what we can do, but also what the broader ecosystem can do. depot is focused on rebuilding the software delivery infrastructure layer for the new throughput we work at. that includes everything from the moment the code is committed to the moment it is deployed. if that sounds exciting to you, we are always hiring, dm me or shoot us an email. if you haven't tried out depot yet and you want to use depot ci or give depot code a try in its very early form, drop me a note and i'm happy to hook you up with some credits.
1
2
291
God damn! Trying to work here.
2
239
Thus is really neat from the folks at @ExtendHQ
AI-generated fake documents grew 500% last year, and humans can no longer tell the difference That's why we’re launching Extend Verify, the first document fraud detection API built for agents Verify catches the fastest growing types of document fraud, including: - fully AI-generated documents - localized AI edits (down to a single modified character) - patched text, overlays, and in-painting from Photoshop, PDF editors, and coding agents - digital tampering after a document was signed - and many more In our internal evals, Verify detected over 96% of AI-generated and tampered documents. Verify is available today: extend.ai/verify/request-acc…
1
3
268
Big update from @devtunehq today: Brand Positioning: see who AI credits with your positioning (you or your competitors) Response Accuracy: flag what AI gets wrong about you, by checking its answers against your own site and docs Then fix things and measure the impact. devtune.ai/blog/ai-positioni…
2
216
DevTune now integrates with our favourite hedgehog @posthog to allow you to track content change impact on AI search referrals and conversions. Pair with our @Cloudflare or edge middleware sensors to get a complete view of AI traffic, and test and guide your content work, docs.devtune.ai/ai-traffic/p…
1
3
180
I can tell a lot about a growth team from their experiment plans. Most plans I see are just ceremony. They tick boxes and make the team feel rigorous. A good plan does the opposite. It forces hard calls before you burn traffic: what you hope to learn, what would stop you early, and whether you even need a test at all. I updated my experiment plan template for 2026 (including a few common traps I still see too often): plg.news/p/how-to-write-a-gr…
2
121
This might well be my favourite ever 404 screen. Nice work @depotdev 🔥
1
1
5
592
This is pretty cool! Opus 🦾
2
223
This is very clever @boristane 👏
2
4
434
I have a LOT of time for what the folks at @trycua are doing. This launch was somewhat overshadowed by the (justified) Jev hype, but it's really great in it's own right, and I'm looking forward to seeing more such models.
1/ Introducing CUA-S1: a family of System One Models, small, specialized, and built for computer use. Today we're open-sourcing CUA-S1-FORMS, the first in the family: github.com/trycua/cua
1
7
724
Love Avec - great app.....and these ads are so refreshing.
Liked The Odyssey? Here’s another Greek epic for you. It Could Have Been An Email Part I: Pheidippides and the Battle of Marathon Directed by Grace Hanna & Derek Manansala Avec is a free iOS app that lets you stay on top of your inbox in just a minute a day. It’s the best way to write an email, whether you just won a war or are running late for your next meeting. Download Avec for free on the App Store. And follow us so you don't miss Part II!
1
2
394
It's honestly amazing to have the ability to solve niggling little pains and frictions that annoyed you or slowed you down, but that in the past the ROI wasn't quite there. github.com/subsy/skill-cabin…
3
318
Ben Williams retweeted
we write awesome blog posts. @lukevmorris fucking cooked with this one. this is a killer in depth look into the story, components, and rationale for building depot ci. also gives a glimpse into how depot ci really just slips into Origin like no other ci provider can do.
1
14
1,295
We keep treating the homepage like the source of truth for AI search. It is not. Across the citation mix we measured for tool-evaluation questions, product and landing pages are a thin slice. Blog and docs carry most of the evidence. Homepage still does positioning. Answer engines are reading one click deeper - whether you meant them to or not. stateofaisearch.dev
1
174
I’ve been using this pattern for a few months now, albeit hand rolled with supporting skills like /dispatch so excited to see this start to be productised with first class support in harnesses. Nice work Cursor team.
Introducing Projects, a new way of working in Cursor. Rather than creating a chat for every task, you work with a coordinator agent in a single, persistent thread. Like @bot, your agent is always on, proactively manages work with subagents, and improves over time.
1
6
1,062
Writers of X !! Just a little reminder about skribe-editor.com - your FOSS markdown AI editing partner. npx skribe-editor@latest ~/draft.md github.com/devtunehq/skribe
1
125