Added Oct 15, 2026
Academic Humanizer
Edits papers and grant proposals: drops stock openers and inflated wording while leaving numbers, citations and claims exactly as written.
- Ecosystem
- Agent Skills
- Skill category
- Writing
- Creator / organization
- AIScientists-Dev
What it is
Academic Humanizer is an open-source academic editing Skill by GitHub user AIScientists-Dev. The team writes many papers and grant proposals and found that AI-assisted drafts kept opening with "In recent years...", inflating claims and running long, while general humanizers, built for blogs and marketing, flattened the precision academic writing depends on. So they added an academic layer on top of blader/humanizer's general rules and calibrated it against the team's accepted papers and funded proposals. It edits wording and voice only: it doesn't generate findings, invent data or change citations, and it isn't designed to evade AI detection. You can also give it your own earlier papers and the target venue so the tone sounds more like you. What sets it apart is that it guards the precision scholarship needs, including for grant proposals such as NSF and NIH.
Who it's for:researchers who want AI help editing papers and grant proposals without losing precision.
Before you start: git clone it into Claude's skills folder, or have Codex or another agent read SKILL.md directly; there's no install script or dependency. There's no upload service or dedicated API key; your paper is processed by the AI you use. The README reminds you that using it doesn't remove your duty to disclose AI assistance under your venue's policy. We suggest calibrating only with your own earlier writing. The license is text the author calls MIT, slightly reworded, so GitHub doesn't recognize it automatically.
Based on the project's public documentation. Not tested by AI Inspiration World.