style-harmonizer

Refine academic prose in Markdown sections by removing repetitive phrasing.

497|38|Updated Jan 7, 2026
One-click install
npx skills add https://github.com/WILLOSCAR/research-units-pipeline-skills --skill style-harmonizer
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: style-harmonizer
Source: https://github.com/WILLOSCAR/research-units-pipeline-skills/tree/main/.codex/skills/style-harmonizer
Command: npx skills add https://github.com/WILLOSCAR/research-units-pipeline-skills --skill style-harmonizer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill addresses the subtle, repetitive "generator voice" or "slot phrases" that can make AI-generated text sound unnatural and unoriginal, even after structural edits.

Core Features & Use Cases

  • Prose Polishing: Removes common stylistic tics like count-based openers ("Two limitations...") and reused discourse stems ("The key point is...").
  • Meaning Preservation: Ensures that factual claims, citation keys, and their anchoring to evidence remain unchanged.
  • Use Case: After a draft is structurally sound but still reads like it was written by a machine, this Skill can be applied to make the language flow more naturally and professionally without altering the core arguments or evidence.

Quick Start

Run the style harmonizer script on the current workspace to refine the prose in flagged sections.

Frequently Asked Questions about style-harmonizer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I remove generator voice and repetitive phrasing from AI-generated academic prose?

To remove generator voice from academic prose, you refine stylistic consistency by targeting repetitive sentence structures and slot phrases, ensuring meaning and citation integrity remain intact. This process eliminates stylistic tics like count-based openers and reused discourse stems in Markdown sections.

What is the best way to polish Markdown sections flagged for style smells without altering factual claims?

The best way to polish flagged Markdown sections is to run a style harmonizer that removes count-based openers and reused discourse stems. It updates the Markdown files directly while explicitly preserving all factual claims, citation keys, and evidence anchors.

Can I use a prose harmonizer on sections identified in a self-loop TODO file?

Yes, you can use a prose harmonizer on sections identified in a self-loop TODO file. The tool specifically operates on flagged sections within `output/WRITER_SELFLOOP_TODO.md` and updates the corresponding files in the `sections/*.md` directory.

Does prose harmonization change citation keys or anchoring to evidence in academic drafts?

No, prose harmonization does not change citation keys or evidence anchoring. The process ensures meaning preservation by focusing exclusively on stylistic consistency and naturalness, leaving factual claims and citation integrity completely unchanged.

When should I apply a style harmonizer to my writing workflow?

You should apply a style harmonizer after your draft is structurally sound but still reads like it was written by a machine. It refines the language flow to sound more natural and professional without altering core arguments or evidence.

Why does my AI-generated text still sound unnatural after structural edits?

AI-generated text sounds unnatural after structural edits due to subtle, repetitive generator voice and slot phrases. A style harmonizer targets these specific style smells, such as reused discourse stems, to make the prose flow naturally.