Rewrites AI-sounding prose to remove AI writing patterns while preserving every claim and citation.
npx skills add https://github.com/BTreeMap/SKILLs --skill humanize-btreemap
Please help me install this Agent Skill. Skill: humanize Source: https://github.com/BTreeMap/SKILLs/tree/main/humanize Command: npx skills add https://github.com/BTreeMap/SKILLs --skill humanize-btreemap
High-intent search queries and answers about installing and using this skill.
Paste the text into the conversation or name a file, and the Skill scans it against 40 documented AI writing tells, then rewrites the flagged passages. Pasted text returns a draft and final rewrite; a named file is edited in place.
The rewrite keeps every claim, number, date, quote, and citation at its original strength and invents nothing. Invariants forbid dropping negations, hedges, or comparatives that would change what a sentence claims.
Yes. Supply a writing sample or personal style file and it outranks the pattern catalog. Deliberate habits like frequent em dashes, personification, or candid asides are kept without asking.
Yes, but reference, technical, legal, and factual text is kept neutral rather than given personality. File mode changes prose only, preserving code blocks, YAML metadata, data, and link targets.
Every flagged pattern also occurs in human writing, so a hit is a style signal, never proof of authorship. Lexical tells count only in clusters, and the calibration reference lists false positives to leave intact.
yuqueAnalyze Yuque documents to extract vocabulary, sentence patterns, and tone into a reusable style profile.
chrisbanesDraft and revise evidence-led publishable prose without invented claims or promotional filler.
my-entourageCross-check factual claims against source documents and output an evidence-status table.
AbdelrhmanUZakiCompresses AI responses into terse fragments while preserving technical accuracy.
