humanizer
Skill2 starsChanged 54 days agoMITView on GitHub
Finds AI writing tells and applies only the rewrites the user confirms, or fixes them in place in text the agent wrote in the same task. Use on docs, comments, PR text, changesets, and any prose before it ships, or when asked whether text sounds generated.
188 other projects use a version of this.
Adopted by
Most recently taken up by
- aditya-vermax/skillstheir own version, first seen Nov 8, 2026
- chrisleekr/skillstheir own version, first seen Nov 8, 2026
- xhqing/CommunityManagerAgenttheir own version, first seen Nov 8, 2026
- CloudChef/atlasclaw-providerstheir own version, first seen Nov 8, 2026
- Fractera/Agent-Engineering-Infrastructuretheir own version, first seen Nov 8, 2026
- Fractera/Agentic-Engineering-Infrastructuretheir own version, first seen Nov 8, 2026
Used with
Projects with it also use economy-mode, personal-voice and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
Microsoft AzureSkilleconomy-mode
Shifts Layer 3 model selection to cost-optimized alternatives when economy mode is active.
Also in 4 projects that use this
petermuirurithuoSkillpersonal-voice
Captures the user's personality, communication style, and values. Use this skill whenever responding to the user, drafting anything on their behalf (emails, CVs, messages, opinions, explanations), or whenever tone, word choice, or how an opinion is delivered matters. Apply it by default for any prompt, not only when the user explicitly asks for "my style" or "my voice.
Also in 3 projects that use this
- Microsoft AzureSkill
architectural-proposals
How to write comprehensive architectural proposals that drive alignment before code is written
Also in 3 projects that use this
- +9 more from Azure
Other versions188 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 188
- NousResearch/hermes-agent★ 262ksame name
- blader/humanizer★ 63ksame name
- QuivrHQ/quivr★ 40ksame name
- The-Vibe-Company/quivr★ 40ksame name
- rommapp/romm★ 14ksame name
- github/gh-aw★ 5.6ksame name
- latitude-dev/latitude-llm★ 4.7ksame name
- github/squad★ 3.4ksame name
- compozy/compozy★ 2.9ksame name
- SRombauts/SQLiteCpp★ 2.8ksame name
178 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 175 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
Check the groups that apply to the text.
- Distinctive: line 29
When you wrote the text yourself in this task, such as a docs page or PR body, fix the tells in place without asking.
“Common” is measured: kept by at least 6 in 10 of the 175 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Humanizer
Make prose read as if a person who knows the subject wrote it.
1. Mark the tells
Check the groups that apply to the text. references/patterns.md has the phrase lists and a before and after for each group. A short text with no match needs no reference.
| Group | Tells | Applies to |
| Language | AI vocabulary, "serves as", negative parallelism, rule of three, elegant variation, false ranges, filler, hedging, generic conclusions | All text |
| Style | dashes or semicolons joining clauses, bold on ordinary words, inline-header lists, title-case headings, emoji | All text |
| Structure | history in the deliverable, internals the reader does not act on, vague referents, remembered numbers, count anchors, unmatched house format, instruction shape | Docs, PR text, tickets, runbooks, skill files |
| Content | undue significance, empty credibility, -ing filler, marketing words, vague attribution, filler sections | Blog and marketing prose |
| Communication | chatbot artifacts, sycophantic openers | Chat output |
2. List, then ask
List every tell without editing (see Output). Then follow the ask skill: one question per
rewrite or closely related group, offering Apply, Skip, or Show more.
When you wrote the text yourself in this task, such as a docs page or PR body, fix the tells in place without asking.
3. Apply what was confirmed
Apply only the confirmed rewrites. Then reread them, ask "what makes this obviously AI generated?", and fix what that turns up.
For code and mixed diffs, the deslop skill runs the same confirm step.
4. Add voice, for blog and marketing prose only
Docs, comments, skill files, and PR text stay neutral. A blog post or marketing page with the tells removed but no voice still reads as generated, so: react to facts with an opinion, vary sentence length, admit complexity ("it works, but it feels like a workaround"), and use specific details and natural asides.
Output
One bullet per tell: path:line, the tell, the rewrite. Put a rewrite longer than a line in a
block under its bullet. A file with no tells gets one bullet saying so. After applying, one
bullet per change and one per skipped tell.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.