AI Humanizer Skills — a measured field survey
68 open-source projects measured against the GitHub API: stars, forks, commit cadence, releases, days idle — plus a blind experiment run on thirteen years of my own writing. Finding: star count has no statistical relationship to maintenance (ρ = +0.197 across 54 prose skills, as measured 2026-07-30). One of the two experiments failed; the post-mortem is published with the data.
This is the work product I’d show a client first — not because it’s a website, but because it’s the method: scripts, snapshots, and conclusions anyone can regenerate.
source: github.com/Daguilar0123/ai-writing-skill-field-guide