humanizer-cli

A CLI wrapper for the blader/humanizer skill. Bundles the humanizer prompt with your text and outputs to stdout — pipe it to any LLM.

No API calls. No dependencies. Just prompt + input → stdout.

Install

npm install -g humanizer-cli

Requires Node.js 18.3+.

Usage

# Pipe a file to Claude
humanizer essay.txt | claude --print

# Pipe stdin, then to llm
cat essay.txt | humanizer | llm

# Read from macOS clipboard
humanizer --clipboard | claude --print

# Just see the bundled prompt (no LLM needed)
humanizer essay.txt

# Output only the system instructions
humanizer --prompt-only

Flags

| Flag | Short | Description | |------|-------|-------------| | --help | -h | Show usage | | --version | -v | Show version | | --clipboard | -c | Read input from macOS clipboard (pbpaste) | | --prompt-only | -p | Output only the system prompt |

Input priority: file argument > --clipboard > stdin > show help.

What it detects

The bundled prompt covers 24 AI writing patterns across four categories:

• Content — significance inflation, notability name-dropping, superficial -ing analyses, promotional language, vague attributions, formulaic challenges sections
• Language — overused AI vocabulary, copula avoidance, negative parallelisms, rule of three, synonym cycling, false ranges
• Style — em dash overuse, boldface overuse, inline-header lists, title case headings, emojis, curly quotes
• Communication — chatbot artifacts, knowledge-cutoff disclaimers, sycophantic tone, filler phrases, excessive hedging, generic conclusions

How it works

The CLI reads your text, prepends the humanizer system prompt, and writes everything to stdout. The output looks like:

[humanizer instructions — 24 pattern definitions, rewrite process, output format]

Humanize the following text:

[your input text]

References

• Wikipedia: Signs of AI writing
• WikiProject AI Cleanup

License

MIT