🛡️ code-health

Instant, zero-config project health scanner with local static analysis and an interactive multi-provider AI architecture review engine—built to run 100% serverless with zero running cost.

code-health is an ultra-fast terminal utility designed for modern developers. It executes a comprehensive repository audit locally in under 30ms, generates a gamified repository score out of 100, maps a remediation roadmap, and optionally spins up an interactive dropdown loop to trigger deep AI code reviews via completely offline models (Ollama) or secure cloud APIs (Google Gemini).

⚡ Key Product Features

• ✅ 100% Serverless & Private: Runs entirely locally on your hard drive. Zero central server routing infrastructure, and zero network data leaks.
• 📦 Static Dependency Audit: Recursively crawls your code workspace strings to cross-reference package.json arrays against actual file imports, instantly identifying dead-weight packages.
• 🖼️ Asset Blob Inspector: Sweeps folders to flag unoptimized static images and media files exceeding 1.5MB that slow down clone metrics and push cloud build costs up.
• 🚀 DevOps Compliance Mapper: Verifies structural base configurations like active unit test suites, baseline documentation files, and standard remote cloud CI pipeline architectures.
• 🤖 Token Batch Splicer (Chunker): Features an advanced text-stream array tokenizer that automatically splits large multi-file codebases into compressed text slices, ensuring long source directories process through LLM context windows smoothly without truncation exceptions.
• 🔄 Multi-Provider AI Compatibility: Pulls installed models dynamically from your system layer via loopback lookups (like smollm2:135m or llama3) or bridges to remote API keys with secure masking prompts.

💻 Technical Stack & Native Modules

The engine operates on a zero-config, ultra-low dependency model to maintain instant shell bootstrapping:

• fs: Low-level synchronous Node file-descriptor streams tracking space weights, file arrays, and parsing strings.
• path: Dynamic operating system path mapping utility ensuring standard compatibility across Windows, macOS, and Linux terminal sessions.
• prompts: Keyboard-controlled cursor selection arrays for menu navigation loops.
• chalk: Direct visual color terminal strings (Green/Yellow/Magenta) formatting diagnostic priorities.

🚀 Installation & Local Environment Setup

Register the tool globally on your system terminal framework straight from your workspace directory:

# 1. Open your core tool folder
cd C:\Users\mprah\code-health

# 2. Lock down node dependencies
npm install

# 3. Bind the application binary shortcut link globally
npm link

🎮 How to Execute

Navigate your shell workspace window to any application project repository on your computer and invoke the main operational trigger:

code-health

📊 Sample Product Workflow Logs

1. Static Scan Analysis Stage (Bypasses UI clears instantly)

====================================================
🛡️  Code-Health Engine Workspace Diagnostics Studio
====================================================
Analyzing directory location: C:\Users\mprah\sync

====================================================
📊  DIAGNOSTIC SCORE: 50/100
====================================================

  ✓ Zero oversized or unoptimized static files found within workspace.
  ✓ All package dependencies matched cleanly against string parse.

  ⚠️ Missing structural layout README.md manual.
  ⚠️ Zero active unit testing architectures found.
  ⚠️ Missing cloud verification deployment hooks (CI Pipelines).

💡 STEP-BY-STEP REMEDIATION TRACK ROADMAP:
  1. Run touch README.md to clarify setup parameters.
  2. Install test beds (jest or pytest) to protect logic layout.
  3. Inject .github/workflows/ci.yml to run cloud tests.

Base metrics scan executed successfully at: 4:29:17 pm
----------------------------------------------------

2. Multi-Provider Interactive Prompt Selection Stage

? Would you like to trigger an advanced AI Architectural Code Review for this project? » (y/N) yes
? Choose your preferred AI Engine backend driver configuration: » Ollama (100% Free, Private, Local Offline AI Engine)

🔎 Interrogating local port for Ollama status... Online!

? Select the local Ollama model to use for analysis: » smollm2:135m (0.25 GB)

🚀 Initializing Source Token Splice Slicing Framework...
Found 15 files. Split into 6 chunks.

Running deep AI code evaluation loop. Please wait...

3. File System Batch Persistence Generation

----------------------------------------------------
⚙️  Dispatching segment payload packet (1/6)... Received Report!
--- 📘 BATCH REVIEW REPORT 1 OF 6 ---
... [Detailed structural security issues & refactored code components displayed on screen] ...
💾 Saved local copy to: ./AI_HEALTH_REPORT_BATCH_1.md

📄 License

Distributed under the open-source MIT License. Completely free to use, modify, or scale worldwide.