humanize-text_skill

This skill removes AI writing tells and makes text sound human by fixing punctuation and tone across documents.
  • Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill humanize-text

  • _meta.json280 B
  • SKILL.md5.9 KB

Overview

This skill humanizes AI-generated text by removing common AI writing patterns and fixing stylistic tells without rewriting the content. It strips em-dashes, filler phrases, structural clichés, and sycophantic openers so the result reads naturally and human-written. Use modes to control how aggressive the cleanup is: Clean, Light, or Preserve.

How this skill works

The skill applies a deterministic set of transformations in order: punctuation fixes, dead-phrase removal, structural changes, formatting adjustments, vocabulary swaps, and tone calibration. For files it can read the input, apply the chosen mode, produce a diff summary of changes, and write or preview the cleaned version. It preserves technical terms, meaning, and the author's intent while removing only AI-specific patterns.

When to use it

  • Cleaning blog posts or articles that feel mechanical or over-polished
  • Polishing emails, pitches, or social posts to sound more natural
  • Preparing documentation or framework text where voice must stay human
  • Reducing sycophantic or formulaic openings in customer-facing copy
  • Before publishing content that must avoid obvious AI provenance

Best practices

  • Start in Clean mode for full de-AI-ing, use Light when you want minimal change
  • Review the diff for files — transformations aim to preserve meaning but may alter punctuation and sentence shapes
  • Keep the author's personality; avoid replacing voice or adding opinions
  • Preserve domain-specific jargon and technical accuracy—skip vocabulary swaps when inappropriate
  • Run a second pass for tone calibration if the original had strong personality

Example use cases

  • Turn an AI-drafted blog post into a natural, varied narrative with contractions and varied sentence length
  • Remove filler phrases and structural clichés from product documentation without changing technical terms
  • Clean up marketing copy that uses puffery and buzzwords, replacing them with plain alternatives
  • Polish customer emails to remove overly formal punctuation and robotic openers
  • Batch-process a wiki or handbook file set and get a diff report of all stylistic fixes

FAQ

No. It avoids altering technical terms, proper nouns, and domain-specific language. Changes focus on style and pattern removal, not content meaning.

Can I control how aggressive the cleanup is?

Yes. Use Clean for full transformation, Light to limit edits to punctuation and dead phrases, or Preserve to make minimal, high-confidence fixes.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational