John Montgomery

16.2K posts

John Montgomery

John Montgomery

@YourDesigncoza

Programming most things AI these days & a Seasoned full time Daytrader ( XAUUSD ) with +20 years experience .

South Africa Katılım Mayıs 2009
353 Takip Edilen681 Takipçiler
John Montgomery
John Montgomery@YourDesigncoza·
@PierrickGT @ryancarson @HelloUntangle @DevinAI Fair question. I don’t see this replacing Playwright. I use Playwright for deterministic, repeatable tests with clear assertions. An AI agent is useful for long messy, changing workflows where you want “real user” exploration Playwright = regression Agent = edge-case smoke test
English
0
0
0
20
Ryan Carson
Ryan Carson@ryancarson·
It costs us ~$33/day to automate a complete e2e test of the new customer onboarding flow for @HelloUntangle. This is a complex process with ~28 unique steps and takes ~99 minutes for @DevinAI to complete it. I get a report every morning with ... 1. PASS/FAIL grade 2. Details of run 3. Videos of Devin completing onboarding About once a week this uncovers an edge case bug we would've missed. Worth every penny. Here's the prompt: ========================= Run the daily E2E smoke test playbook. Create a fresh account on untangle.us, complete onboarding with Grace, submit case details, and test all 14 Grace chat tools. Record the session and report results. Only interact with trusted data. ========================= This calls a playbook (like a skill) called "Daily E2E Smoke Test — Signup, Onboarding, Case Details & Grace Tools" that tells Devin how to do this. This playbook is ~200 lines of markdown. This includes spinning up a new temp @agentmail inbox so we can sign up on prod with a fresh account. The playbook is versioned and we regularly improve it.
Ryan Carson tweet media
English
23
16
186
18.1K
John Montgomery
John Montgomery@YourDesigncoza·
@browomo This is all BS : “I am Claude Code, Anthropic’s official command-line tool. I can help you complete various software engineering tasks, such as: Writing and modifying code Debugging and fixing bugs Searching and exploring codebases Running commands and tests Creating PRs and ..
English
0
0
0
188
Blaze
Blaze@browomo·
This Chinese guy created agents in Claude Code for landing pages and single-handedly serves 47 small businesses a month, taking $400 from each. He built a system of 7 agents on Claude Sonnet 4.6 that analyzes Google Maps in small towns, finds small businesses without websites there, and over 1 weekend takes each one to a finished mockup with video and cold message. No assistant, no sales team, no SDR. Just him, a MacBook, an iPhone, and 1 API key. And traditional web design agencies keep teams of 8 people on salary for the same order flow, while his expenses are only tokens and subscriptions to Lovable, Higgsfield, and Calendly. 7 agents work through 1 orchestrator on Claude Code Router. Usage is about 3 million tokens a day, the average API bill is about $480 a month. All 7 go through MCP servers and write shared state to the file system, without shared state in memory and without race conditions, and 1 of them lives right in the iPhone and picks up positive replies from the subway, a taxi, or on walks. And here is the system prompt he put into the orchestrator before launch: "You are the orchestrator of a solo agency that sells ready-made websites to local businesses. You delegate read-only tasks to 6 sub-agents and own all writes. sub-agents: // Scout (walks through Google Maps in selected cities, looks for narrow niches: 5+ years on the map, fewer than 50 reviews, no website or a website from 2014, but high ratings) // Diagnoser (for each lead writes a 50-word diagnosis, hero angle, tone matched to the industry, and a cold message under 70 words) // Builder (generates a landing page mockup in Lovable through MCP only for the top 5 leads per day, with the sharpest diagnoses and the biggest gap) // Filmer (pulls 5 screenshots of the mockup and through Higgsfield renders a 10-second vertical video 1080x1920 with a soft zoom) // Pitcher (sends a personalized cold message through the right channel for the niche: email to roofers, SMS to tradesmen, IG DM to salons, LinkedIn to realtors) // Checker (runs every message through evals for personalization, absence of AI markers and buzzwords before sending) // Mobile (lives in the iPhone, handles positive replies in real time, books Zoom calls in Calendly through MCP while the owner is on the go). You never let 2 sub-agents touch 1 lead. You stop and request approval from the human only when a deal exceeds $3,000 or the reply rate in a niche for the day drops below 12%." Meaning the system knows what it is and within what boundaries it is allowed to act. It knows it is supposed to find leads on its own. It knows it is supposed to take each one to a mockup, video, and cold message without intervention. It knows the human only steps in when a deal goes above $3,000 or the reply rate stops converging. → The system runs 24 hours a day → Scout goes through about 220 local businesses on Google Maps per day and leaves 30 new leads in the queue → Diagnoser outputs 30 structured diagnoses + briefs + cold messages per day → Builder assembles 3 to 5 finished landing pages in Lovable for the sharpest leads → Filmer renders a 10-second vertical video in Higgsfield for each one → Pitcher sends 30 personalized messages per day across 4 channels with a reply rate of about 14% → Checker runs every message through evals before sending And only when a deal breaks $3,000 or the reply rate for the day drops below 12% does the orchestrator wake the owner. And when the owner at that moment is sitting in the subway or a taxi, the Mobile agent in his iPhone picks up 1 move on its own: replies to a fresh positive reply from a dentist, books a Zoom through Calendly synced to the local time of the client, and puts the lead back in the queue. The owner only has to tap "approve" and in just 10 minutes join the call. Here is what the system writes in his log during 1 of the Saturdays: "scout report: 218 businesses checked in Austin, Denver, and Miami, 34 without a website, 19 with a website from 2014, 6 with an active redesign request in reviews. passing top 30 to diagnoser." "pitcher: 30 cold messages sent across 4 channels, 14 replies, 5 positive, 3 Zoom calls booked for Sunday. passing to closer." "builder: landing page for Westside Cosmetic Dentistry built in Lovable, 5 sections, mobile, soft beige. URL placed at /Users/dev/maps-agency/clients/westside/v1. filmer launching Higgsfield." "eval flag: deal with The Lotus Salon at $3,400 exceeds the approved limit of $3,000. sending for manual review." He has no server of his own and no separate backend. Just a local file sandbox at /Users/dev/maps-agency, an MCP router, 1 API key to Claude, and the same key forwarded to Claude Code on his iPhone. Out of everything I have seen this year, this is the cleanest one-person agency for selling websites to small businesses: $480 a month on the API, about $18,800 into the account, and between them 7 prompts, 1 file system, and 1 phone in the pocket.
timbidefi@timbidefi

x.com/i/article/2051…

English
230
1.1K
9.9K
2.7M
John Montgomery
John Montgomery@YourDesigncoza·
@krassenstein She's nothing .... Besides, he's got a country to go blow apart no time for this hag in front of him :::
English
0
0
1
8
Brian Krassenstein
Brian Krassenstein@krassenstein·
Trump is so disrespectful. He literally cut in front of Queen Camilla while she was shaking hands. Every day is another embarrassment for our country.
English
13.3K
22.4K
96.2K
3.7M
John Montgomery retweetledi
Jason
Jason@mytechceoo·
CEO obsessed with token maxxing
English
282
1K
13K
1.9M
John Montgomery
John Montgomery@YourDesigncoza·
@codewithpri Linux has a learning curve, for sure : But if you're a professional you should not mind the learning curve, once you see the advantages ... you can never go back to the BLUE & FROZEN days :::
English
0
0
0
135
Priyanka Lakhara
Priyanka Lakhara@codewithpri·
Linux is only free if your time has no value.
Priyanka Lakhara tweet media
English
417
68
1.4K
248.8K
Lesedi
Lesedi@LifebeLifing_SA·
@ChrisExcel102 Last time a white man came to Africa, wanting to help black people well we all know how that ended
English
7
6
303
6.4K
ChrisExcel
ChrisExcel@ChrisExcel102·
Starlink cost R950 a month A local Internet Fiber network cost R130 a months with fast internet Why does Elon Musk thinks we desperately need that thing here 😭😭😭
ChrisExcel tweet media
English
928
3.2K
21.2K
452.6K
John Montgomery
John Montgomery@YourDesigncoza·
@Ernest866459 @Real_RobN When I do leave one day ... NEVER babba NEVER ! Interesting that you work on systems developed and built by the so called oppressors , By the way do you even know what this means "Digital sovereignty advocate" ... big words for a small brain !
English
1
0
0
6
Tommy Robinson 🇬🇧
Tommy Robinson 🇬🇧@TRobinsonNewEra·
President Trump: "We suspended all refugee resettlement, except for persecuted South Africans.. ...There's a very horrible thing going on in South Africa, it's a genocide, they kill people if they're white!"
English
548
5.1K
29K
279.6K
John Montgomery
John Montgomery@YourDesigncoza·
@Ernest866459 @Real_RobN More that that has left the country in just the last 2 years ::: What's your point ? We are still NET loosing investment ... you are a Special Kinda Stupid!
English
1
0
0
10
SkhokhoIT
SkhokhoIT@Ernest866459·
@Real_RobN Yet in South Africa 🇿🇦 👇🏽👇🏽👇🏽
SkhokhoIT tweet media
English
17
30
383
15.8K
Jackie Phamotse
Jackie Phamotse@JackiePhamotse·
Dr. Naledi Pandor exemplifies exceptional leadership, inspiring admiration for her unwavering dedication to public service, astute intellect, and steadfast commitment to promoting South Africa's global interests. Her poised and generous leadership style renders her an exemplary figure, worthy of emulation.
English
10
9
62
3.3K
Garry Tan
Garry Tan@garrytan·
I got inspired by Karpathy’s LLM Wiki, implemented it in my OpenClaw and then extended it with my own skills and a full Postgres pgvector implementation. Want to be one of the first to try GBrain? This is my personal opinionated version of Karpathy’s LLM Wiki on OpenClaw
Garry Tan tweet mediaGarry Tan tweet media
English
106
95
1.2K
115.8K
John Montgomery
John Montgomery@YourDesigncoza·
@Gabbar0099 Must say seems fair ... What did they expect will happen when you poke the bear !
English
0
0
0
4
Gabbar
Gabbar@Gabbar0099·
One of the most widely shared videos worldwide 💔🇵🇸 Israel did this.
English
5.7K
19.9K
52.9K
3.7M
John Montgomery
John Montgomery@YourDesigncoza·
@JackiePhamotse BlaH BlaH .... Oom Jannie is always to blame : By the way the books you wrote that you say are being used in schools looks like a load of 💩 : "a controversial adult fiction novel ..." Yip teach this SHIT to your generation & see how far you get !
English
0
0
1
31
Jackie Phamotse
Jackie Phamotse@JackiePhamotse·
End of an era: Jan van Riebeeck will no longer dominate school history books! This is truly the beginning of a better generation. We are so sick of colonialism taking over our education systems. Our own people need to be studied, our own books need to be used in our schools.
English
352
370
1.3K
29K
John Montgomery
John Montgomery@YourDesigncoza·
@NickSpisak_ @karpathy @steipete @tobi Been using Obsidian as a living wiki for my dev projects for many moons, not just notes, but actual architecture decisions, functions, edge cases, and reasoning. Building or debug, I RAG my own knowledge base. My code + my decisions = the context. AI updates MY "WIKI/Context"
John Montgomery tweet media
English
2
0
4
634
Nick Spisak
Nick Spisak@NickSpisak_·
Made an updated version this weekend Here's how you do it (raw notes) > Grab @karpathy's latest gist (in the first comment) > Download @steipete summarize CLI > Download yt-dlp > Download obsidian > Download @tobi qmd --> Setup a node or Golang CLI called "brain" --> Have it index all your youtube data, AI agent data (jsonl files) --> Get your X data by requesting an archive in your settings --> Setup vaults for each domain/topic area --> Ask questions with your agent and qmd
Nick Spisak@NickSpisak_

x.com/i/article/2040…

English
72
327
3.6K
593.9K