47thtechcorner/raycodes_humanizer
Developer toolsこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
Humanizer Skill 3.0: Make AI Writing Sound 100% Human Locally! - Local AI prose humanizer skill for Ollama, Claude Code, and Skills CLI
概要
Humanizer transforms dense, AI-generated technical documentation into clean, conversational developer blog posts. It strips away robotic fluff and corporate jargon while keeping all your code blocks, command lines, and links completely intact. Run this single PowerShell command to install dependencies, create the Modelfile, and register noema2b in Ollama: Alternatively, pull and use llama3.2:3b via Ollama: Execute the main Python engine to convert your input document: The transformed, human-readable blog post or developer conversation will be created inside outputs/outputs.md along with exact recorded generation time. - Python 3.12 (Native AST Markdown Parser) - Ollama (noema2b:latest / llama3.2:3b) - Markdown (inputs/input_doc.md) - Markdown (outputs/outputs.md) 1. Clean up AI-drafted CHANGELOGs so developers get direct updates without corporate filler. 2. Turn dry architecture specifications into engaging tutorials for engineering teams. 3.
README
以下は原文 README の抜粋です。原文を見る。
⚡ Humanizer Content Transformer Engine
Humanizer transforms dense, AI-generated technical documentation into clean, conversational developer blog posts. It strips away robotic fluff and corporate jargon while keeping all your code blocks, command lines, and links completely intact.
| Step 1: Skill Install | Step 2: Model Registration | Step 3: Run Engine |
|---|---|---|
| Install Humanizer via CLI / PowerShell | Register noema2b or pull llama3.2:3b |
Execute python main.py |
🛠️ Installation & Setup
1. Skill Installation Options
Option A: Skills CLI Installation (Recommended)
npx skills add blader/humanizer --global
Option B: PowerShell Direct Skill Download
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/blader/humanizer/main/SKILL.md" -OutFile "SKILL.md"
Option C: Claude Code Plugin Installation
/plugin marketplace add blader/humanizer
/plugin install humanizer@humanizer
2. Model Download & Registration Options
Option A: Noema 2B GGUF Model (Recommended)
Run this single PowerShell command to install dependencies, create the Modelfile, and register noema2b in Ollama:
pip install urllib3 ; echo "FROM hf.co/NoemaAI-labs/Noema-2B-GGUF`nPARAMETER temperature 0.7`nPARAMETER presence_penalty 1.5" > Modelfile ; ollama create noema2b -f Modelfile
Option B: Llama 3.2 3B Model
Alternatively, pull and use llama3.2:3b via Ollama:
ollama pull llama3.2:3b
3. Run Command
Execute the main Python engine to convert your input document:
python main.py
The transformed, human-readable blog post or developer conversation will be created inside outputs/outputs.md along with exact recorded generation time.
💻 Tech Stack
- Core Logic: Python 3.12 (Native AST Markdown Parser)
- Primary AI Backend: Ollama (
noema2b:latest/llama3.2:3b) - Input Document: Markdown (
inputs/input_doc.md) - Output Result: Markdown (
outputs/outputs.md)
📁 Repository Structure
humanizer/
├── inputs/
│ └── input_doc.md
├── Modelfile
├── main.py
└── README.md
💡 5 Real-World Use Cases
- Open-Source Release Notes: Clean up AI-drafted CHANGELOGs so developers get direct updates without corporate filler.
- Technical Developer Blogs: Turn dry architecture specifications into engaging tutorials for engineering teams.
- Repository READMEs: Strip robotic phrasing from project landing pages to boost maintainer credibility.
- API Quickstarts: Rewrite formal endpoint guides into beginner-friendly walkthroughs.
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add 47thtechcorner/raycodes_humanizer