myai - Firewall-Safe AI CLI Tool

A production-ready CLI tool that works on restricted college networks by using HTTPS-only communication (port 443) with a disguised backend API.

Features

✅ Firewall-Safe: Works on heavily restricted Wi-Fi (HTTPS port 443 only)
✅ No Local AI: All processing happens on your secure backend
✅ Smart File Extension Detection: Auto-formats output based on file type
✅ Production-Ready: Built with security and reliability in mind
✅ Easy to Use: Simple command-line interface

Installation

npm install -g myai-cli

Quick Start

Basic Usage

# Print to terminal
myai "explain binary search"

# Save to file with automatic formatting
myai "write a DFS algorithm in C++" --out dfs.cpp
myai "explain paging in OS" --out paging.md
myai "create a Python REST API" --out api.py

Examples

Generate C++ Code

myai "implement quicksort with comments" --out quicksort.cpp

Generate Python Script

myai "create a web scraper using requests" --out scraper.py

Create Markdown Notes

myai "explain TCP vs UDP" --out networking.md

Generate JSON Data

myai "create a sample user schema with 5 users" --out users.json

Generate Configuration

myai "create a docker-compose for nginx and postgres" --out docker-compose.yml

Supported File Types

The CLI automatically detects file extensions and formats output appropriately:

Code Files (clean code only, no markdown):

• C++: .cpp, .c, .h
• Python: .py
• JavaScript/TypeScript: .js, .ts, .jsx, .tsx
• Java: .java
• Go: .go
• Rust: .rs
• And many more...

Documentation:

• Markdown: .md (formatted with headers, code blocks, etc.)
• Text: .txt (plain text)

Data Formats (valid syntax only):

• JSON: .json
• YAML: .yaml, .yml
• XML: .xml

Configuration

Custom Backend URL

Set a custom backend URL via environment variable:

export MYAI_BACKEND=https://your-backend.vercel.app/api/query

Or on Windows (PowerShell):

$env:MYAI_BACKEND="https://your-backend.vercel.app/api/query"

Command-Line Options

Usage: myai "<prompt>" [--out <file>]

Options:
  --out, -o    Output file path (with extension)              [string]
  -h, --help   Show help                                      [boolean]
  -v, --version Show version number                           [boolean]

Examples:
  myai "write a DFS algorithm in C++" --out dfs.cpp
  myai "explain virtual memory" --out notes.md
  myai "create a REST API in Python"

Why This Works on College Wi-Fi

1. HTTPS Only (Port 443): Uses standard web traffic that firewalls allow
2. No WebSockets: Only simple HTTP POST requests
3. No Streaming: Single request-response pattern
4. Disguised Backend: API looks like a generic query service
5. No Local AI: Doesn't download models or run suspicious processes

Troubleshooting

"Cannot connect to backend"

1. Check your internet connection
2. Verify the backend URL is correct
3. Try accessing the backend URL in a browser
4. Check if your firewall blocks HTTPS requests

"Backend error (401)"

The backend API key is not configured. Contact the backend administrator.

"Rate limit exceeded"

Too many requests. Wait a minute and try again.

Security

• ✅ All communication over HTTPS (encrypted)
• ✅ No API keys stored in CLI
• ✅ No telemetry or tracking
• ✅ Open source and auditable

License

MIT

Support

For issues or questions, open an issue on GitHub.