Building and shipping useful apps. Sharing experiments, lessons, and real numbers in public.
Joined February 2026
- Tweets1.8K
- Following1.2K
- Followers1.3K
- Likes3.2K
Pinned Tweet
Building and shipping useful apps in public.
Sharing the real work:
- ideas worth testing
- launches
- what users do
- numbers
- mistakes along the way
Current projects: Mobile apps, Linklee.click, and Chalkwork.
Follow if you are building useful things too.
This is super exciting, excited to see computer use and upcoming features too!
Tristan retweeted
DeepSeek V4.1 Flash is live in Command Code.
6x usage on $10/mo GOAT for a week.
$60 usage. 154K requests. 7.8B tokens.
GOAT plan is the best AI coding plan.
Available in all plans and API.
🐐
Tristan retweeted
Your ChatGPT Work agent can now use websites that require you to sign in.
Take over the cloud browser to log in, then let your agent continue the task. Your login persists across sessions, so you only have to sign in once.
Hey!
Looking to meet other people working on:
🚀 Tech
🛠️ SaaS
🧑💻 Indie Hacking
🧠 AI Tools
🪽 Hermes
📱 Product Dev
📈 Marketing
What are you currently working on?
Let’s connect 👇🏻
One rule for the apps I build next:
If it needs a sign-up, AI, a dashboard, or ten screens before it helps, it is too much for v1.
Payment can happen through the App Store or Google Play. No extra product account.
Useful first. Complexity later.
Hey builders, the weekend is here! ☀️
What are you building right now?
Drop it below - SaaS, AI agents, apps, automations, digital products, anything.
Tell us what it does and who it is for.
Let us get some eyes on good work and meet cool people building. 👇
Tristan retweeted
GPT 5.6 Sol just hit CursorBench.
The economics are brutal for Anthropic.
Fable 5 Max: 70.5% at $17.32 per task. 103,525 tokens.
GPT 5.6 Sol Max: 67.2% at $5.22 per task. 28,320 tokens.
95% of the performance. A third of the cost. 73% fewer tokens.
Fable 5 kept the crown on raw score. It lost on everything that shows up on your invoice.
And remember: GPT 5.6 comes with limits you can actually live with.
The frontier is not an intelligence war anymore.
It is a value war.
AI tools are most useful to me when they make the boring loops easier:
- draft the post
- compare the positioning
- summarize objections
- turn replies into product notes
- make tomorrow's plan sharper
Still human judgement. Just less blank-page friction.
Question for people trying to grow on X:
Would you use a small browser extension that scores posts by reply opportunity?
Not an auto-reply tool.
More like:
"this post still has momentum"
"this one is probably dead"
"this author/topic matches your lane"
Useful or noise?
I am also thinking about an X growth tool.
The pain:
you can write a good reply under the wrong post and get nothing.
The useful version would not auto-reply.
It would help you spot which posts are still worth replying to.
If you have a small project, product, newsletter, or service:
what is the one action you want visitors to take after opening your profile link?
I got one clear Linklee signal yesterday:
"weak analytics"
If you use a link-in-bio page, what do you actually wish it showed you?
1. clicks per block
2. traffic source
3. email signups
4. product clicks
5. booking/support interest
6. something else
If you have a link in bio:
what’s the most annoying part of your setup?
1. too many tools
2. newsletter/signup is awkward
3. weak analytics
4. selling/support links feel messy
5. booking links feel bolted on
6. it feels too serious for what you need
I’m validating Linklee this week.
A simple home for your small useful things:
- links
- newsletter/signup
- products
- booking/support
If you make small products, apps, guides, or hobby projects, I’d love blunt feedback.
Yesterday’s feedback sharpened Linklee:
It may not be for people already “building a company.”
It may be for people making small useful things who need a simple home for links, interest, products, newsletter/signup, and booking/support.
Before they need a full site.