CodeSage AI

AI-powered code reviews inside VS Code. Get instant, expert-level feedback on bugs, security, performance, and code quality -- powered by DeepSeek via HuggingFace.

Features

AI Code Reviews -- Analyze any file or selection and get structured, actionable feedback with severity ratings, line references, and code fix suggestions.
Inline Diagnostics -- Issues appear as native VS Code squiggly underlines. Click the lightbulb to apply AI-suggested fixes in one click.
CodeLens Integration -- "Review" buttons appear above every function, method, and class. Click to review a single function in isolation.
Review Profiles -- Four specialized modes tailor the AI's focus:
- General -- Balanced review: bugs, security, performance, readability
- Security -- OWASP-focused audit: injection, secrets, auth, cryptography
- Performance -- Bottleneck analysis: complexity, memory, I/O, caching
- Clean Code -- Craftsmanship: SOLID, DRY, naming, nesting, refactoring
Streaming Responses -- Reviews render in real-time as the AI generates them, so you see results immediately.
Multi-Language -- Works with Python, JavaScript, TypeScript, C++, Java, Go, Rust, PHP, and any language with VS Code symbol support.
Secure -- API keys stored in VS Code's encrypted SecretStorage. Never written to files.

Installation

From Source

git clone https://github.com/Abdus-Sami01/codesage-ai.git
cd codesage-ai/extension
npm install
npm run build

Press F5 in VS Code to launch the Extension Development Host.

Prerequisites

VS Code 1.97.0 or higher
Python 3.8+ with huggingface_hub installed:
```
pip install huggingface_hub
```
HuggingFace API Key -- Get one at https://huggingface.co/settings/tokens

Quick Start

Open the Command Palette (Ctrl+Shift+P)
Run CodeSage: Set API Key and paste your HuggingFace token
Open any code file
Press Ctrl+Shift+R or run CodeSage: Review Code
View the review in the panel that opens beside your editor

Issues will also appear as inline diagnostics (squiggly underlines) in the editor. Hover over them and click the lightbulb to apply fixes.

Settings

Setting	Default	Description
`codesage-ai.model`	`deepseek-ai/DeepSeek-R1`	AI model for reviews
`codesage-ai.maxTokens`	`4096`	Maximum response length
`codesage-ai.temperature`	`0.3`	Response creativity (0 to 1.5)
`codesage-ai.pythonPath`	`python`	Path to Python interpreter
`codesage-ai.reviewProfile`	`general`	Review focus: general, security, performance, clean-code
`codesage-ai.enableCodeLens`	`true`	Show Review buttons above functions
`codesage-ai.enableStreaming`	`true`	Stream results in real-time

Commands

Command	Shortcut	Description
CodeSage: Review Code	`Ctrl+Shift+R`	Review the active file or selection
CodeSage: Review Function	--	Review a single function (via CodeLens)
CodeSage: Set API Key	--	Store your HuggingFace API key securely
CodeSage: Select Review Profile	--	Switch between review profiles

Architecture

VS Code Extension (TypeScript)
  |
  |-- reviewCode.ts / reviewFunction.ts  (commands)
  |-- reviewService.ts                    (spawns Python subprocess)
  |-- reviewPanel.ts                      (Webview rendering)
  |-- diagnosticsProvider.ts              (inline squiggly lines)
  |-- codeLensProvider.ts                 (Review buttons on functions)
  |-- statusBar.ts                        (profile indicator)
  |
  v
code_review.py (Python backend)
  |
  |-- Reads JSON from stdin
  |-- Calls HuggingFace API (DeepSeek model)
  |-- Writes JSON to stdout (streaming or batch)

Development

npm install          # Install dependencies
npm run watch        # Build + watch for changes
npm run build        # One-time build
npm run build:prod   # Production build (minified)
npm run lint         # Run ESLint
npm run package      # Create .vsix package

Contributing

Fork the repository
Create a feature branch: git checkout -b feature/my-feature
Commit your changes: git commit -m "Add my feature"
Push to the branch: git push origin feature/my-feature
Open a Pull Request

License

MIT

CodeSage AI(DeepSeek)

Samiullah Atta