.png&w=384&q=75)
Will Leatherman
Founder at Catalyst Content
AI personalization at scale is a mail merge with more compute. It scrapes a name, writes a clever first line, and ships the same body to everyone. Buyers spot it instantly, and a wrong-fact email replies below a plain one.
This is the exact cold-email engine we built live in Claude Code. It starts from one real signal, infers the pain behind it, grades how much you actually know, and writes an email under 120 words with the AI tells stripped out. When the signal is thin, it hedges instead of faking.
It's the engine we built live, packaged so you can install it and run it on your own signals in about 15 minutes.
What's in the download:
The full Claude Code project, with a README and build steps that include every prompt
A rules file that bans the two AI tells (the forced bridge and the invented specific)
A brief step that grades your evidence as strong or thin before a word is written
A writer that obeys the rules and hedges when it should
Two real, sourced example signals and a sample weekly draft output
Built by Will Leatherman and the team at Catalyst, who've built content and outbound systems for 150+ B2B companies. Every line the engine writes traces back to something true, so you can send it without wincing.
Free
Turn one real signal into a cold email worth a reply, with the AI tells stripped out. Built in Claude Code.