Everyone can now spot machine text on sight, the em-dashes, the "you're absolutely right", the invented technical terms nobody at any job ever used. I write a lot with agents, and I did not want my published words to smell like that. So I built hyphos, a small open tool that learns how I actually write, from my own words, and rewrites AI drafts to sound like me. This is the build log, including the parts that went wrong.
The corpus came first. My raw material was every message I've ever typed into my coding agents, plus a decade of sent email. Extracting "what I typed" turned out to be the real work. The first pass over my agent transcripts produced 1.27 million words, which sounded great until I measured where they lived, 91 percent sat in messages too long for anyone to have typed. Pasted logs, quoted docs, other people's text. Nobody types 1500-word prompts. So the filtering is mechanical and merciless, machine-shaped lines stripped out of every message, lengths capped at what a human plausibly types, and any message carrying a machine tell inside supposedly-mine text thrown away whole, 606,000 words in that last cut alone. I am left with 80,470 words that behave like typing, and that number I trust.
