Skip to content
SkillModel-led3,811 tokens in use

academic-humanizer

Edits AI-assisted academic drafts and NSF or NIH proposals for clarity and the author's voice, ties every claim to evidence, and leaves numbers and citations untouched.

5,711 stars on the repository Custom / UnknownAIScientists-Dev/academic-humanizer
git clone https://github.com/AIScientists-Dev/academic-humanizer ~/.claude/skills/academic-humanizer
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing academic-humanizer costs

108tokens idle
3,811tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

writes filesdeclared

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter15/15
  • Structure25/25
  • Examples6/15
  • Output15/15
  • Size15/15
  • Resources0/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Needs attention
  • Development activity8
  • Maintenance0
  • Community44
  • Maturity24
  • Momentum40

academic-humanizer is an editing pass for papers, theses, rebuttals and funding proposals. It removes the marks of generic AI drafting, such as inflated significance, empty intensifiers, formulaic openers and clause-stacked sentences, while keeping the calibrated hedging, passive voice and technical terms that scholarship needs.

It checks that each empirical claim has a number, figure or citation and that no verb is stronger than its evidence. A separate proposal mode keeps the ambition reviewers expect and concentrates effort on the opening pages. It is not meant to hide AI use, and disclosure rules still apply.

What You Get

  • An audit listing each detected pattern, its location and the proposed fix
  • A rewrite with the same structure, every number, equation and citation unchanged
  • A change report covering patterns removed, claims softened or given evidence pointers, and voice notes
  • Proposal-specific fixes for vague importance, method-as-aim, dependent aims and boilerplate broader impacts

Common Use Cases

  • Cleaning the abstract and introduction of a machine learning paper drafted with AI help
  • Rewriting an NIH Specific Aims page so the gap, hypothesis, aims and payoff are clear on page one
  • Tightening an NSF Project Summary under Overview, Intellectual Merit and Broader Impacts
  • Matching a draft to the voice of an author’s earlier papers and the register of a target venue

What Makes It Unique

General humanizers target blogs and flatten scholarly precision. This skill keeps evidence-tied hedging, passive voice, the plural we and every symbol, and adds claim-to-evidence checks so a verb never outruns the data. Its proposal mode treats vision plus feasibility as the goal: it keeps bold framing when a credible plan backs it and flags missing support for the author to supply.

Under The Hood

The skill is a single instruction file with before and after examples and no scripts. It is allowed to read, search, write and edit files in your project, so it can open your manuscript or LaTeX source and save the rewritten text. It does not fetch web pages, call APIs or run commands. Its rules forbid inventing numbers, citations, funding or partners.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers