tomwils1967 retweeted
GitHub is a goldmine for designers.
I found 10 insane repos that turn your designs into real code, some with Claude and AI. All free.
Bookmark this for later:
1. Penpot
The open-source design platform, a free Figma alternative you can self-host. It has an MCP server, so AI can read and work on your designs.
github.com/penpot/penpot
2. Framelink MCP for Figma
Give Cursor and other AI coding tools your Figma files. They one-shot designs far better than when you paste screenshots. 16K+ stars.
github.com/GLips/Figma-Conte…
3. Onlook
A canvas where designers build real websites and prototypes. Connect the Claude Code or Codex account you already have and describe what you want.
github.com/onlook-dev/onlook
4. Frontend Design skill
Anthropic's own Claude skill for UI that doesn't look like a template. Bold choices on palette, type and layout.
github.com/anthropics/skills
5. Screenshot to Code
Turn a screenshot, mockup, Figma design or screen recording into clean HTML, React or Vue code with AI.
github.com/abi/screenshot-to…
6. Excalidraw
A hand-drawn style whiteboard for wireframes and flows. Collaborative and end-to-end encrypted.
github.com/excalidraw/excali…
7. tldraw
Build your own infinite canvas app in React, multiplayer included.
github.com/tldraw/tldraw
8. shadcn/ui
Beautifully designed components you copy into your project and make your own.
github.com/shadcn-ui/ui
9. Lucide
A free, consistent open-source icon library in SVG.
github.com/lucide-icons/luci…
10. Lottie
Airbnb's library that plays After Effects animations natively on web and mobile. Design it once, ship it without an engineer rebuilding it.
github.com/airbnb/lottie-web
What did I miss?
tomwils1967 retweeted
Claude is f*cking insane for building a personal finance system...
Send this prompt to Claude and it will turn every receipt, invoice and bank export into one monthly money report you can actually read
Then go even further with the full setup below:
tomwils1967 retweeted
We all know how this one is going to go.
Shockingly light sentence because it was a black guy experiencing mental distress, and the judge will agree the intent was theft, not murder.
🚨Newly released 911 audio and surveillance footage expose the terrifying final moments of 68 year old Frank Imhoff, who was killed while simply loading his trailer at a Port Neches, Texas, storage facility.
According to investigators, the Nederland resident was working outside his pickup truck when a suspect identified as Dekovin Flanagan Jr. suddenly leaped across the hood of the vehicle and stole it. As the truck sped off, Imhoff was run over, suffering fatal injuries.
Flanagan was later tracked down and apprehended by law enforcement at a nearby apartment complex. He is currently being held on a $1 million bond facing murder charges.
Investigators are still piecing together a motive, noting that it remains entirely unclear why Imhoff's truck was targeted or what triggered the violent escalation.
Case Backstory & Investigation Details
The fatal incident occurred on August 16, 2026, at Central Maxi Storage located on Ridgewood Drive in Port Neches, Texas.
The Victim: 68 year old Frank Cameron Imhoff of Nederland, Texas, was a dedicated father and grandfather affectionately known as "Campa" to his grandchildren.
Family members noted that he was hardworking and would naturally try to protect his livelihood.
According to the Port Neches Police Department and investigative affidavits, the suspect identified as Dekovin Keith Flanagan Jr. (21, of Port Arthur, Texas) had initially wandered into and rummaged through an unoccupied storage unit on the property.
Flanagan then sprinted out, hopped the hood of Imhoff’s pickup truck while it was being loaded with a trailer, forced his way into the driver's seat, and attempted to speed away.
As Flanagan accelerated in the stolen vehicle, surveillance footage captured Imhoff running alongside the driver's side trying to stop the theft before being knocked down and run over by his own truck and trailer.
Following a chaotic morning that involved erratic behavior and property destruction at a nearby apartment complex, law enforcement tracked Flanagan down to the Ridgewood Apartments, where officers found him hiding in a closet.
Dekovin Keith Flanagan Jr.’s specific upcoming trial date or docket appearance has not yet been publicly scheduled or published in local court records.
#FrankImhoff
#PortNeches
#PortNechesTX
#SoutheastTexas
#TexasNews
#InvestigativeJournalism
#JusticeForFrankImhoff
tomwils1967 retweeted
Gm
Paper Americans leach off of legacy Americans
Would love to see the military service rate of Somalis and Indians vs some random town in the rust belt
No idea, but would guess @RoKhanna ‘s family doesn’t have much of a military service history in the US
Prove me wrong
tomwils1967 retweeted
I just killed SEMRush (Local), BrightLocal and Whitespark.
These PATHETIC rank trackers charge $50 to $300 a month to resell you Google Maps data that costs about a dime per scan.
So I built LocalSEO.to.
You plug in your own DataForSEO key and pay the raw data cost. No monthly fee. No credits that expire.
Here's what it does:
→ Geo-grid heatmaps up to 9×9, for as many keywords as you need
→ Scheduled rescans, with a timeline to flip back through every past scan
→ Compares any two scans and shows every point that moved
→ Ranks you against competitors by share of the top 3 across the grid
→ Tracks the Local Pack next to your Maps rank
→ Sends each client one live report link with your logo, marked verified
→ Audits a Google Business Profile and its website, 0 to 100, with fixes in priority order (beta)
→ Finds every listing with the wrong name, address or phone (beta)
→ Ships with an API and MCP server, so Claude Code can run all of it
The math for 10 clients, 3 keywords each, a 7×7 grid weekly:
😂 Semrush Local: $3,600 a year
😂 BrightLocal: $747 a year
😂 Whitespark: $600 a year
🫦😮💨🫨🤤😩 Your own key: about $46 a year in data
Pay once. Lifetime access. 14-day refund.
The price goes up $10 every 5 sales, and the pay-once deal ends Dec 16.
📌 Get it here:
localseo.to
TrustMRR:
trustmrr.com/startup/localse…
i'm going to make $200k in the next 30 days and clap every indiehacker cheeks in public
follow along by signing up for the newsletter:
jackychou.com/
here's how
tomwils1967 retweeted
Sister.
Wife just showed me this, Holy shit 😂
This was amazing! Absolutely dead 😂 🤣
The ending really makes it, unfortunately that’s me.
tomwils1967 retweeted
Grok Bot token saver is here!
(your weekly limit finally lasts the whole week)
We hire Grok Bots for every job now, but each bot still drags old chat, re-reads the same documents and runs the big model on easy steps, which empties your limit by Thursday.
So I just built Volt, one Grok Bot whose only job is to make every other bot you own last the whole week.
You hire it once, and it checks every bot you have for 7 leaks.
Every leak gets a fix, and nothing changes until you say OK. Volt then gives you:
- one fix list, heaviest bot first
- a fix on each bot after your OK
- a one-line report for every bot
- a 3D Power Grid that shows where your limit goes
Here's the Volt template: inside the AI Profit Boardroom
(all 3 Volt messages are copy-paste ⭐ )
Volt hunts the 7 leaks hiding in the bots you already run, like:
- old chat
- the same documents
- the big model on easy steps
- long answers + giant pastes
- busy routines + double work
With Volt, your bots do the work, not the thing burning through your week.
Your bots still think like any Grok Bot.
However, in this case, Volt owns the cleanup, so your bots stop paying for the same old chat every single run.
Here's why it matters, from the builder I posted about, who checked a full week of usage line by line:
- 93% of their tokens went on rereading old chat history
- one heavy bot dropped from up to 94K tokens a step to 44K
That's about half the tokens on one bot, and most of the waste came from old chat nobody was reading.
You can keep building new bots while you test. Once a bot starts eating your limit, hand it to Volt.
This answers the real questions, like:
- Which bot is eating my weekly limit?
- Which steps actually need the big model?
- Is a routine running more than it needs to?
Once it's hired, Volt fixes each bot after you say OK, with a one-line report for every fix.
So you're no longer paying for the same old chat on every run.
Instead, you spend your limit only on the work that actually needs it.
The clip below shows Volt's 3D Power Grid, where cables carry power to every bot and every leak lights up.
If you want to see where your Grok Bot limit really goes, I posted a whole breakdown of how one builder fixed it for around 60 bots.
It covers the routines that found nothing, the free doorbell that replaced 720 bot runs with about 2, and the 5 team-room rules.
Read it below.
Want Volt + the system ? Get them inside the AI Profit Boardroom 👉 skool.com/ai-profit-lab-7462…
This Grok Bot usage fix is honestly kind of wild.
93% of one builder's tokens went on rereading old chat history, not on answers.
Around 60 Grok Bots hit the weekly limit in under 3 days, so they checked a full week of usage line by line.
Here's what cut it:
→ Turned off routines that found nothing (43 of 45 watchdogs found nothing)
→ A free Cloudflare "doorbell" checks a page every hour with no AI, so one job goes from about 720 bot runs to about 2
→ Reset the heaviest chats from a 600-word handoff file (one bot dropped from up to 94K tokens a step to 44K)
→ Banned "got it", "thanks" and "okay" between bots in team rooms
Nothing gets deleted. The old bot just gets hidden.
The biggest mistake: they moved to a plan more than 3x bigger with the same habits. It ran out in under a day.
Fix the habits first.
Then decide if you need more.
To start this week:
→ Export your usage and find your 3 heaviest bots
→ Turn off the routines that find nothing
→ Paste the 5 team-room rules into every room
→ Give your heaviest bot a fresh start, one bot at a time
tomwils1967 retweeted
Your favorite wedding photo deserves more than a spot in your camera roll. 🤍 Turn the moment you said “I do” into a personalized Oil Painting, finished with beautiful raised brushstrokes you can actually see and feel.
tomwils1967 retweeted
Validate your AI startup idea (and collect cash) in 7 days:
Day 1: Pick the idea
Look for work someone is already paying a human to do:
-Job posts and Upwork VA roles
-Agency RFPs
-Compliance work nobody owns (OSHA logs, EPA reports, license renewals)
-Reviews that say "I wish it had X" on Reddit, G2, and Capterra
-Websites with manual contact forms, "call for a quote", RFQs
Have Claude Code scrape the Reddit threads and reviews, then rank the pains by how often they show up.
Day 2: Pick the wedge
The wedge is one small job that gets you in the door. Pick it by answering 4 questions:
1) Who is the one person my solution benefits? Not "the company." Karen in accounting, Lucy in HR, Doug in sales.
2) What does that person do by hand every week? Pull it from your Day 1 list.
3) Is each of their customers worth $5K-$10K+? Then one saved deal pays for you.
4) Do you already know the industry? Your network and experience are a huge head start.
Then fill in one line:
"I make [person] look like a superhero by [taking over the job] so they [outcome]."
Example wedge: make Doug in inside sales look like a superhero. Every quote request gets researched, enriched, and answered in minutes instead of the 42-hour B2B average.
Days 3-4: Build the hook
-Use Paper to mock up 20 landing pages with different headlines for different pain points
-Pick the one that hits, build it in Claude Design or Claude Code, deploy on Vercel
-Record a short VSL in Riverside
-Record a quick Loom as proof you're a human
A Loom angle I've seen work for folks: "I filled out your lead form. It took 43 hours to hear back. No, I'm not a real lead, but I can help your business."
Then build your lead list:
-Exa Websets or Firecrawl
-Filter to find leads that have a contact us or request a quote form
-Enrich to find CEO name, LinkedIn, open job posts, recent news, CRM they use, content they write
-Feed it all back into Claude to write custom messaging per prospect
25 companies x 4-5 contacts each = 100-125 people to reach.
Day 5: Talk to customers (MOST IMPORTANT STEP BY FAR)
-Use a hooky subject line like "your contact form is broken" (lowercase)
-Or just their first name. Literally "corey" would be the subject line. Gets insane open rates.
-Give value to earn the next step. Something like "Send me a company name, I'll research the lead for free. All I need is the URL."
-If your outreach takes longer to read than to reply, it's too long
-Record every call in Fathom/Granola so Claude can read it after
Day 6: Build the MVP offer
Run it through Hormozi's value equation: push dream outcome and likelihood of success up, push time and effort down.
@jrdnmix's version he shared on my pod:
-Offer: every inbound lead lands enriched
-Reverse offer: "Let's run your last 50 leads and see what you missed"
-Starts Monday, first briefing in week 1, ROI in week 2
-All they give you: inbox access and 30 minutes
-Price: $2,000 for the inbound audit
For the ongoing service, price against the SALARY:
$120K salary x 1.34 = ~$161K fully burdened = ~$13,400/month. I'd pitch it at ~$8,000/month.
Day 7: Pitch it and read the signal
-3+ paid on your entry offer = winner
-1 paid = you have something
-Pushback = pivot the offer and run it again
-No replies = fix the messaging or move on
"I'd buy that" is not a commit. A commit is ONLY cash collected.
You won't walk out of sprint one with a $10K/month business. Plan on a couple. The average founder pivots 4+ times (if not more).
Hand this entire post to Claude Code or your Grok Bot and have it run Days 1-4 with you.
@jrdnmix broke this playbook down on a recent episode of the Build With AI podcast with me.
If you want the link to the full episode lmk. It was a banger.
tomwils1967 retweeted
Just a normal classical liberal out here talking about doing classical liberal things at the classical liberal conference.
tomwils1967 retweeted
Yeah France is gone. Don't know what else to say.
“ACAB” scandé en chœur et fièrement au Bataclan, au concert d’Artemas, à la veille du 13 Novembre, c’est ignoble 😤 !
Il y a 11 ans, dans cette même salle, un commissaire et son frère d’armes 🇫🇷 sont entrés en sachant qu’ils risquaient de ne pas ressortir en vie, vous comprenez ça ?
Ils ont risqué leur peau et neutralisé l’un des terroristes islamistes, puis les policiers de la BRI-PP et du RAID sont intervenus pour libérer les otages en se faisant tirer dessus.
Alors, à ceux qui ont crié “tous les flics sont des bâtards”, regardez bien la salle où vous dansiez : ici , des gens ont cru qu’ils ne reverraient jamais leurs proches et des flics sont entrés pour les sauver, avec des familles qui, elles aussi, attendaient qu’ils rentrent.
Vous avez insulté les flics dans la salle même où ils ont risqué leur vie pour en sauver d’autres.
Vos propos me font vomir, c’est minable !
tomwils1967 retweeted
Over 1,000 Afghan “refugees” have been resettled in Utah under Spencer Cox (since FY 2021).
Over 1,000 Congolese.
Over 2,300 unaccompanied alien minors.
No telling how many family members they’ve all pulled in via family visas.
My paternal (Cox) and maternal (Irons) ancestors were born here more than 475 years ago (and like every good LDS member I have receipts 😂). Does that make me more American than Bo? You can oppose the insane Biden open borders and mass immigration (I do and did), support removing criminals (I do), support a limited number of regulated work/education permits (I do) without believing that someone who swore the naturalization oath last week is less American than someone whose ancestors came here 400 years ago.🇺🇸🦅
tomwils1967 retweeted
The Parasites of Peace
To all the little priests of international law, trembling over words about men with swords, you’ve grown fat on a peace you neither earned nor understand.
Your courts, treaties, conventions, and solemn declarations were built in the shadow of men who carried weapons. Men who stood in the cold, buried their friends, and killed when killing actually became necessary. Their blood bought the quiet in which you learned to argue about the morality of their existence.
The lawyer is a luxury of civilization. The warrior is a necessity of the human condition. One can exist in a burning world. The other needs someone to put out the fire before he can open his briefcase.
Yet somewhere along the way, the lawyer began confusing his ability to interpret the law with the power to make the world obey it. He mistook the courtroom for the battlefield and his signature for a sword.
Now he lectures the men who keep him safe, armed with nothing but paper and the certainty that someone else will bleed when the paper fails.
And you call this civilization?
There’s a particular arrogance in demanding that the sword answer to men who’ve never felt its weight, especially when those men mistake their advice for command.
You can write a thousand resolutions condemning the wolf. He’ll still come through the gate. And when he does, you won’t send a lawyer to meet him. You’ll send the same hard men you’ve spent your comfortable lives trying to tame.
Then you’ll discover what your peace was worth, and who was paying for it.
Law can govern the sword. It can’t replace it. And a civilization that forgets the difference will eventually learn it from men who have no intention of obeying either.
tomwils1967 retweeted
That face you make when you spent your tenure calling bridges and roads racist instead of fixing things
Pete Buttigieg takes a swipe at the Trump administration after Friday's FAA ground stop threw the nation's air travel into chaos, joking he'd personally help stranded travelers rebook their flights from a hotel bar in Denver.
The former transportation secretary, at the "What's Next" conference hosted by Majority Democrats, questioned why the aviation system's built-in safeguards failed to prevent an outage of that scale. "The system has many redundancies in it so that normally if there's one point of failure, something else can catch it," he said. "It is not clear to me why that redundancy did not prevent an outage of this magnitude."
The FAA lifted the ground stop after more than four hours, but flights remained delayed and canceled over the weekend.
Full story here foxnews.com/politics/buttigi…
tomwils1967 retweeted
Yes, getting flown to America and enrolled in a ton of welfare programs is just like the 1620s.
Just gets more brilliant every time I see it.
tomwils1967 retweeted
Update on yesterday’s FAA ground stop. We now know what happened, and it’s worse than “old equipment.”
It wasn’t a cyberattack and it wasn’t SMART. The FAA says there’s zero evidence of either.
Here’s what happened. L3Harris crews were replacing an end-of-life copper line with fiber at one remote FAA site. A router involved in that work started flooding the network with more than 1,000 connection attempts every four to five seconds. Radar feeds and radio links to aircraft became unstable. After about 45 minutes of nobody knowing why, the FAA grounded the entire country.
Once they isolated that one remote site, everything came back.
So let me get this straight. One router at one site took down air traffic control for all 50 states. The FAA has now disabled the faulty function on roughly 60 routers of the same type nationwide. It still doesn’t know whether the router was defective or whether it was installed or configured wrong.
Oh, and this has happened before. In 2009, a Harris technician misconfigured a router while moving the same network to fiber and disrupted 129 air traffic facilities. Same contractor, same migration, same failure mode, 17 years apart.
This is exactly the point from my last post. Legacy modernization doesn’t fail because of the shiny new system. It fails at the seams, in the routine change nobody thought could take down the whole country. Good change management means one site can’t flood the network, isolation happens in minutes instead of hours, and nobody runs a copper-to-fiber cutover on critical infrastructure without a tested rollback and a hard limit on how far a failure can spread.
None of that is new. We learned it in 2009. The question is why we had to learn it again.
Tech is so fun...
Transportation Secretary Sean Duffy said FAA contractor L3Harris was working at a remote location to switch telecommunication lines from cooper to fiber, which led to the longest nationwide shutdown of flights since the September 11, 2001, attacks reut.rs/3Vm3wKO
Duffy unsure if equipment or contractor error caused shutdown - News
tomwils1967 retweeted
This is f*cking insane. Anthropic CEO Dario Amodei was asked what a kid finishing high school should do now. His answer:
"AI can do entry-level law work. It can do entry-level finance. It can do entry-level consulting
We can't deny that it will also eliminate jobs, and probably a large number of them"
In a few minutes he names which jobs go first and the one skill that keeps you in the game
His own prediction: 50 percent of entry-level white-collar jobs disrupted in the next 1 to 5 years
Prompts → Agents → AI Coworkers → Entry-Level Work Done By AI
Farming to factories took generations. This time AI hits every step of a career at once
His advice fits in one line: "Of course, learn to use AI"
The people who start this year won't be competing with AI, they'll be the ones running it
Everything below is free. Start today:
Anthropic - Prompt Engineering Interactive Tutorial: the fastest way to learn to direct a model
github.com/anthropics/prompt…
Anthropic Courses: from your first prompt to running agents
github.com/anthropics/course…
Anthropic Cookbook: real task recipes you can copy today
github.com/anthropics/anthro…
Microsoft - Generative AI for Beginners: 21 lessons, no degree required
github.com/microsoft/generat…
Microsoft - AI Agents for Beginners: build agents that do the entry-level work for you
github.com/microsoft/ai-agen…
Hugging Face - Agents Course: free and certified, start to finish
github.com/huggingface/agent…
Google Cloud - Generative AI: Google's own hands-on notebooks
github.com/GoogleCloudPlatfo…
OpenAI Cookbook: code for every common task
github.com/openai/openai-coo…
DAIR.AI - Prompt Engineering Guide: the reference everyone links to
github.com/dair-ai/Prompt-En…
Andrej Karpathy - Neural Networks Zero to Hero: understand what is under the hood
github.com/karpathy/nn-zero-…
CrewAI: run a small team of AI workers
github.com/crewAIInc/crewAI
Microsoft - AutoGen: put several agents on one job
github.com/microsoft/autogen
Awesome LLM Apps: hundreds of real builds to learn from
github.com/Shubhamsaboo/awes…
DeepLearning.AI - Short Courses: free 1-hour courses from the people who teach this
deeplearning.ai/short-course…
Total cost: $0
You don't need another year of tuition to learn this. You need to start telling AI what to do
95% will graduate and compete with AI for the same entry-level job
The 5% who learn to run it this year will manage the agents doing that job
Save this and send it to someone graduating this year









































