JevGrade – CLI that scores text against custom rulesets

Reproducible, scriptable text quality scoring for writing and publishing workflows

Problem

Makers are shipping CLI tools that grade text against custom rulesets, and products that let 10,000 AI readers react to a post before publishing. This shows writers and content teams want to turn subjective judgment into repeatable automated checks, but existing tools are either dev-only or social-simulation heavy, leaving a gap for a lightweight, configurable scorer that plugs into CI or writing workflows.

Solution

A CLI plus config file where users define rules (banned cliches, required data points, readability thresholds, brand tone). The tool scores text and reports each hit with fix suggestions. Integrate with Git hooks or CI to gate blog posts, docs, and marketing copy before publishing.

Users

Technical writers, content marketing teams, documentation engineers, indie bloggers

MVP

YAML rule syntax covering regex matches, length/readability thresholds, and LLM semantic rules; CLI batch-scores files and directories; JSON and human-readable output; Git pre-commit example; a starter ruleset for common copy issues.

Revenue

Free open-source CLI; subscription for cloud rule sync and team sharing; per-repo or usage billing for CI scenarios; one-off sales of industry rule packs.

Why now

The flood of AI-generated content makes human-sounding, information-dense writing scarce, so teams need automated QA instead of manual reads; a CLI is light, embeds easily, and an indie dev can validate quickly.