FralRater

Offline readability scoring for 8 reader personas, with a lost-reader map and rule-based rewrites. Zero deps, zero network, zero AI.

Visit website

FreeOpen SourceOtherContent Writing1 impressionslaunches May 211 comment

>log in to comment

fralsare[maker]· 20d ago
I built FralRater as a completely offline desktop app that scores your writing for specific readers, maps where they get lost, and rewrites the hard parts—without any network or AI. It’s MIT‑licensed and ready to use right now. I’d love to hear what you think!

FralRater is a desktop app that answers three readability questions the usual tools skip: too hard *for whom*, *where exactly*, and *what do I change*. - **Score** — Flesch-Kincaid grade & ease, Gunning fog, SMOG, jargon count + per-100-word density, long-sentence and passive-voice ratios, hard-word percentage. - **Audience** — fit scores for 8 reader personas (child, general public, professional, executive, patient, developer, legal, academic) with per-paragraph confidence bars. - **Lost reader map** — per-sentence difficulty curve; one point where comprehension drops. - **Rewrite** — deterministic rule-based rewrites (passive→active, buried-lead fix, jargon swaps, sentence splitting) with a sentence-by-sentence diff and before/after grade cards. - **Compare** — two-version radar chart + head-to-head metric table. Engineering constraints I set for it: - the app logic has **zero npm dependencies** — plain ES modules + browser APIs, no build step - **100% offline**: no network calls, no telemetry, no account - the packaged Electron app serves its own UI from an in-process static server in the main process and locks it down with a strict CSP - CI (GitHub Actions) builds all five artifacts: AppImage, deb, rpm, NSIS installer, portable zip MIT-licensed. Text never leaves the machine. Downloads: https://github.com/fralsare/FralRater/releases

Open Source Other Content Writing

FralRater is an offline desktop app that scores readability for specific reader personas, maps where readers get lost, and suggests rule-based rewrites.

for: Writers, editors, and technical communicators who need persona-specific readability analysis without internet or AI dependencies.
pricing: open source
license: MIT

fralsare/FralRater 0 0JavaScriptupdated 20 days ago

v0.1.1FralRater 0.1.120 days ago

Key features

8 features of FralRater

Score — Calculates Flesch-Kincaid, Gunning fog, SMOG, jargon density, long-sentence and passive-voice ratios, and hard-word percentage.
Audience Fit — Provides 0-100 fit scores for eight reader personas with per-paragraph confidence bars.
Lost Reader Map — Shows a per-sentence difficulty curve to pinpoint where comprehension drops.
Rewrite Engine — Deterministic rule-based rewrites (passive→active, jargon swaps, sentence splitting) with before/after diffs.
Compare Mode — Radar chart and head-to-head metric table for two document versions.
Rhythm Heatmap — Visualizes sentence-length variance to highlight monotony or choppiness.
Voice Note Input — Allows dictation via Chrome/Chromium Web Speech API.
Zero Dependencies — Pure ES modules, no npm packages, and 100% offline operation.

Use cases

Assess a legal brief for executive-level readability and get actionable rewrites.
Compare a draft and its edited version to see readability improvements across personas.
Identify which sentences in patient instructions cause comprehension loss and fix them.
Analyze technical documentation for developers to reduce jargon and passive voice.