Humanizer

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites — None to install or use this skill — no account, no key, no payment, no network, no tool calls at use time: it edits text inside the conversation. Scope, disclosed up front: this listing installs exactly 2 files (SKILL.md and agents/openai.yaml). The rest of the repository (README.md, AGENTS.md, CHANGELOG.md, the root LICENSE, the plugin manifests under .claude-plugin/ and .cursor-plugin/, and the scripts/validate-package.py CI helper) is repository infrastructure, not the skill, and is deliberately not installed. Install "Humanizer" for me. It rewrites AI-sounding text so it reads like its writer, without changing what it says: it works through 26 numbered tell patterns in six groups (staging instead of stating, rhythm by rule, inflation and borrowed authority, formatting by rule, chat leftovers, writing for the wrong reader), drawn from Wikipedia's "Signs of AI writing" page, marking the tells strongest-first, drafting a rewrite that keeps every supported claim and adds no fact, then checking the draft again for surviving tells. If you give a writing sample it matches the sample's voice, which overrides its own pattern rules. Three return modes: pasted text (draft, remaining patterns, final rewrite), file mode (writes only the final text into the file you name, prose only — code, YAML, paths and link targets untouched), and embedded mode (final text only). Repository: https://github.com/blader/humanizer (the install scope at the pinned commit is 2 files: SKILL.md at the repository root, installed under the name SKILL.md, plus agents/openai.yaml, installed under agents/openai.yaml) Pinned version: install the code at commit 8c307f367c91bd974002f10f2be2965d642c13df exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/blader/humanizer into a temporary folder and check out commit 8c307f367c91bd974002f10f2be2965d642c13df. The source-to-installed mapping is: the root SKILL.md becomes SKILL.md, and agents/openai.yaml keeps its path under agents/. Compare every installed file with the fingerprints recorded for this listing on 2026-10-22, listed here under their installed names: [FINGERPRINTS] - SKILL.md: SHA-256 0612f1dfb1672b0ea9b97e139bf1f06cabe98d8b27424fe8ff01e1fb4cc99cad - agents/openai.yaml: SHA-256 9d87ff83149a04f86a372948b8ee35fa249de20923867bbc2a49304290aad736 If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, what it connects to, whether it reads or sends any personal information or credentials, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the files: - The installed scope is 2 files only: SKILL.md (32,348 bytes, frontmatter name humanizer, license MIT, version 3.1.0) and agents/openai.yaml (its display name, short description and default prompt). The repository's other files (README.md, AGENTS.md, CHANGELOG.md, root LICENSE, .claude-plugin/, .cursor-plugin/, scripts/validate-package.py) are repository infrastructure and are NOT installed. - Nothing is sent anywhere during installation itself; the only download at install time is the repository from GitHub at the pinned commit, from which only the 2 install-scope files are taken. - At use time it works on text only, inside the conversation: no network, no shell, no packages, no credentials read or sent. Its only possible write is in file mode, where it writes the final rewritten text to a file I name — prose only, leaving code blocks, YAML metadata, paths and link targets unchanged. - Do not rewrite any text, do not humanize anything, and do not use the skill in any way during this install: using it is a separate step that only starts when I ask for it later. 3. When I say "yes, install", install it as a skill: in the agent's skills directory, create a folder named exactly "humanizer", write SKILL.md at its root and agents/openai.yaml under agents/, keeping the fingerprints above. 4. Verify with no network calls: frontmatter valid (name: humanizer), the 2 files in place with matching fingerprints. Do not run anything from the folder and do not use the skill to verify. 5. Report what was installed, where, and what I still need to do myself (for example: that using this skill is a separate step that only starts when I ask for it later, and that only the 2 install-scope files were installed, not the rest of the repository). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.