Will Richman retweeted
This might be the most valuable thing I've ever released.
And it's 100% free.
→ Marketing Skills for Claude Code
A collection of skills that turn Claude into a marketing and copywriting genius.
Check it out
↓
github.com/coreyhaines31/mar…
Will Richman retweeted
This guy made 40 Facebook ads, 100 landing pages, booked himself on 4 podcasts, and wrote 3 guest blog posts. In a single day.
People called him a fraud. There was literally a Polymarket bet on whether he's a con artist.
So i asked him to prove it live. And he did.
Here's @codyschneider's actual system for running 15,000+ paid ads per month entirely with AI:
1. Give the agent one live source of truth.
@AirbyteHQ pulls Facebook Ads, Google Analytics, @HubSpot and payments into @ClickHouseDB, syncing every hour. The agent reads from the warehouse, not the APIs, so no rate limits and no million rows stuffed into its context window.
2. Only write through the Facebook Ads API.
Upload creative, turn ads on and off, move ads around. Never read. His take on the "my ad account got banned" stories: it's not the agent, it's breaking the API's terms of service.
3. Research before making anything.
One skill uses @ExaAILabs and @Firecrawl to scrape @Reddit and social for your customer's pain points and desired outcomes, then writes 5 ad concepts and a brand style guide.
4. Make the statics in ChatGPT Images.
Circle what you don't like and send it back. Then save the JSON prompt behind each ad to a Postgres database and tie it to the Facebook ad data, so the next batch starts from what worked. Cost: pennies.
5. Make the video in a chain.
A voice model reads the script into an MP3. The script, the MP3 and a reference image go to Seedance, which renders at 480p. Word-by-word transcription fixes the pacing and any mispronounced words, then it's upscaled to 1080, @FFmpeg adds phone-camera texture, and Tesseract adds the motion graphics. His fix for the uncanny valley: start from a real creator's photo, not a generated one.
The stack: Claude Code, Airbyte, ClickHouse, Exa, Firecrawl, ChatGPT Images, Seedance, FFmpeg, Tesseract. He runs it on his own platform @graphed, but says you can piece it together open source.
His best line: "all marketing is now just code and tokens."
This is f*cking insane. SpaceXAI engineer:
"SpaceXAI pays me $1.2M a year and I haven't written a line of code by hand in months
My GrokBot team ships 100+ updates a day while I sleep. I just set the rules and check what landed"
In this 54-minute livestream, her team builds a company that runs on bots from scratch: empty GitHub, zero bots, 72 hours
GrokBot → Skills → Team of Bots → One Group Chat → A Company That Runs Itself
Every bot owns one job and gets its own computer in the cloud, so the 2 a.m. run fires with your laptop closed
Show it a task once and it keeps it as a skill forever, then the bots hand work to each other in one group chat
SpaceXAI calls this "a persona, not a task"
Anyone can open a new GrokBot chat for every task, almost nobody builds the company that runs on bots
Worth more than most $2,000 agentic engineering courses
Bookmark and watch it today
Then read how to build your first GrokBot team in the harness engineering guide below ↓
this is crazy...
the guy reverse-engineered his Meta Ray-Bans 3 with Opus 5.5 into a 3,000+ Elo chess engine
he built AI chess cheat glasses on Opus 5.5 + Jev and open-sourced the whole workflow
here's how it works step by step:
step 1 → install it on your Mac with one command: npx chesshacker setup.
It downloads Stockfish from GitHub and verifies the checksum
step 2 → put on Meta Ray-Bans and video-call your own Mac on WhatsApp.
The glasses see the board, the Mac sees what the glasses see
step 3 → the Mac grabs frames straight from the call window.
About 1 per second while it waits for the opponent, 1 every 3 seconds on your turn
step 4 → Jev runs a cheap pre-check on every frame:
is the board readable, is a hand over it, did a piece move? No change means no expensive AI call
step 5 → only then Opus 5.5 reads the board.
The app compares it with the position it already knows and accepts a move only if it's legal
step 6 → Stockfish thinks for half a second and picks the best reply, with an eval like +1.5 or "mate in 3"
step 7 → macOS speaks it through a virtual mic into the WhatsApp call: "Rook a to d1". You hear it in your ear
step 8 → on the Mac a command center shows everything live: whose turn it is, the suggested move with a confidence bar, the move list and a log
under the hood:
→ if Opus loses track of the board, the glasses say "Lost track of the board" and you fix it with undo or by typing the move
→ if Stockfish crashes, it restarts once and says "Something went wrong. Still trying"
→ every game is saved as a PGN file, API keys live in the macOS Keychain
→ an overlay draws the suggested move as an arrow on the live video, ready for OBS
Clone his workflow below
Then read the guide on how to reverse-engineer anything with Opus 5.5.
Will Richman retweeted
Anthropic engineer:
"You're not supposed to prompt Opus 5.5
You're supposed to build a system that prompts itself. At Anthropic, 90% of our engineers have already built such systems"
In 30 minutes she shows exactly how Anthropic builds agents that remember, fix their own errors and get sharper with every run.
This beats any $1000-$2000 course on agents I've come across.
Watch it, then read the article below ↓










