What it is
docfriction generates friction logs for online documentation pages, walking through each section and asking TypeSafe Jev a fixed set of typed questions. It is for developer advocates and writers who want to find where readers get stuck.
How it uses Jev
For each section, Jev answers one Choice question over ten friction types plus no_friction and other, one Score question over four reader outcomes for severity, and several Noul (yes/no) checks for prerequisites, expected result, undefined terms, code/prose match, and unexplained placeholders. Probabilities become findings via thresholds; uncertain answers are marked 'needs review'.
Primitives:choicescorenoul
Technique worth stealing
Use typed questions (Choice, Score, Noul) with probabilities and thresholds to turn subjective documentation friction into deterministic, CI-gateable findings.
Try it
uv tool install docfriction; docfriction https://docs.example.com/quickstart --dry-run; export TYPESAFE_API_KEY=...; docfriction https://docs.example.com/quickstart
Evidence
Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.
- Jev-centricyes0.92
- Shows a System One patternyes0.65
- Handles uncertaintyno0.04
- Measuredno0.04
- Runnableyes0.96
- Worth recommendingno0.21
- Model replicano0.04
- Problem scopescore on a 0–2 scale1.25
- About Jevyes0.97
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.