Stop putting "$10k MRR" in your bio if you’re too scared to sync your Stripe. 👨🍳🔥
Twitter is becoming a gallery of fake screenshots and "trust me bro" numbers. It’s killing the soul of building in public.
So, I built MakerCookS from my dorm in Freiburg. It’s a minimalist portfolio with one ruthless rule: You cannot type your MRR.
It ONLY syncs via Stripe API. 100% verified. 100% honest. Real builders don't hide. Claim your kitchen link below. 👇
https://t.co/CUgNgA5y6A
Get paid to wait
The Claude Code spinner might be the most watched line on Earth.
So I turned it into an ad marketplace.
Advertisers bid on it. You keep 50% of the money.
Install the extension → get cash from ads.
Introducing Kickbacks
I spent time testing Playabl UGC games and building a small browser-based Game Loop Evaluator.
Built a quick game loop evaluator for it.
It opens a public playable link, auto playtest, analyzes what the player sees, then scores the game loop and suggests creator-ready fixes.
The next big social platform won't start with photos or videos. It will start with games.
https://t.co/EOEBWDUx2D is the TikTok for user-generated games: anyone can play, create, publish, and monetize.
In 5 days, they hit 1M organic plays across 3,000 games.
Hey, did anyone know any Discord server / Telegram group that I can discuss MAS or Agentic workflow with other people? Having some practical problem and want to discuss with some one rather than AI(already asked).
@psdrbhy_Lu Hey I 'm also trying to learn building multi-agent system?
One question: do you write all your code or at least some of it? or just architecting the structure and let coding agent do the work?
已經看完你的文章了,推測是前者
Building ScriptedAgent for a little dugen game🎮.
👩🔧Pipeline now:
RunnerAgent -> ObserverAgent -> ReplayAgent -> DetectorAgent -> ReporterAgent -> ReviewerAgent
Trying to integrate Llama3 into ReporterAgent,the other just scripted.
✅Using fixture issue to test it first.
Week1 build log
Start learning AgentXGaming. Start by not making game but how to build steady, replayable environment.
✅Understand MDA framwork
✅Understand every case can be broken down:
Mechanic > Agent behavior > Observed dynamic > QA risk > Evidence > Detector
B2B software is not a great model right now.
And I predict a small shift towards services/agency model (in B2B).
Just talked to a founder who proved it. For 2+ years, he spent $100k building an AI SaaS and only made $150,000. His margins evaporated into paid ads & compute costs.
So he killed the saas and pivoted to a services model. $1-2k per month per client for [redacted] services.
He's doing $1M/year now (and got there quickly).
His margins? 70%!! Because most of the agency work it's not being done by humans, but rather AI agents and automations.
Meanwhile a lot of SaaS founders I talk to are reinvesting everything back into the business in the name of growth, and are lucky to be seeing even 30% margins!!
Also:
- SaaS multiples are in the toilet.
- And DIY tools like Claude Code quietly killing growth
This is just something I noticed as I talk to founders all day.
This is mostly anecdotal, but I'm seeing less fast growing B2B saas products lately and more consumer apps and agency/services businesses crushing it.
I'm struggling with the IRS 990-PF data pipeline for my agent.
Trying to fetch raw XML from IRS S3 to automate grant intelligence (deadlines, history, etc.). I’m using ProPublica’s API as the index, but it's missing data —even for giants like the Ford Foundation.
@tibo_maker It's This is a great story! Perhaps you could consider including a detailed overview of the real costs and the technology stack alongside the complete narrative in one central profilo?
First time getting feedback from cold eamil outreach.
Also first time actually go through entire mvp -> validation, and it's wonderful. 🤩
Really appreciate the first ten customers. You guys are awesome.
First time getting feedback from cold eamil outreach.
Also first time actually go through entire mvp -> validation, and it's wonderful. 🤩
Really appreciate the first ten customers. You guys are awesome.
So base on my study, if you are a fresh builder just like me, builders now use:
- Umami for light project and keep data to yourself
- PostHog for further analysis like Session Replay and A/B tests
- Baremetrics for MRR, Churn Rate, LTV
PostHog is a general best choice.