🔍 CodeSpecto

AI-powered Project Intelligence for Developers

CodeSpecto is an advanced AI code assistant designed for developers to analyze, debug, optimize, review, and secure their code projects. Built using Streamlit, LangChain, and Google Gemini AI, it provides actionable insights, code suggestions, and security recommendations in an interactive and stateful interface.

🚀 Features

• Multi-Mode Analysis

  • General Guide: Get guidance and suggestions on your project.
  • Debugger: Identify bugs in code files automatically.
  • Code Optimizer: Receive optimization suggestions for better performance and readability.
  • Review Your Code: AI-based code review with detailed suggestions.
  • Security Scanner: Detect potential security vulnerabilities and suggest fixes.

• File-Aware Context

  • Analyze multiple files at once.
  • Preserve context for stateful discussions per file.
  • Ask follow-up questions about specific files without losing previous analysis.

• Diff & Patch Suggestions

  • Highlight suggested changes.
  • Apply AI-generated patches directly.

• Interactive Chat

  • Stateful chat interface with AI.
  • Optional queries after initial analysis, context preserved.

• Beautiful & Responsive UI

  • Enhanced color-coded UI.
  • Loading spinners and status messages during analysis.
  • Expander sections for detailed AI insights.

• Supported File Types

  • Python (.py), JavaScript (.js), TypeScript (.ts, .tsx, .jsx)
  • Java (.java), Go (.go), JSON (.json), Markdown (.md)

📂 Architecture & File Structure

CodeSpecto/

• app.py # Main Streamlit interface and UI
• langchain_client.py # LangChain + Google Gemini AI integration
• requirements.txt # Python dependencies
• core/
  • file_indexer.py # Index and manage uploaded files
  • code_parser.py # Parse code and extract snippets
  • context_builder.py # Build AI context per file
  • diff_engine.py # Extract/apply AI suggested diffs
  • security_scanner.py # Scan files for security issues
  • README.md # Project documentation

⚙️ Installation

1. Clone the repository:

git clone https://github.com/yourusername/CodeSpecto.git
cd CodeSpecto

2. Create a virtual environment and activate:

python -m venv .venv
source .venv/bin/activate   # Linux/macOS
.venv\Scripts\activate      # Windows

3. Install dependencies

pip install -r requirements.txt

4. Create .env file and add your Google Gemini API key:

GOOGLE_API_KEY=your_api_key_here

Running the project

streamlit run app.py

• Upload your project files in the sidebar.

• Select the mode (Debugger, Code Optimizer, Security Scanner, etc.).

• Click "Analyse Project".

• View AI analysis, suggestions, and optional patch diffs.

• Use the optional chat box to ask follow-up questions.

🧠 How It Works (Technical Overview)

1. File Indexing

• Uploaded files are read and stored in session state.
• Files are parsed into snippets for AI processing.

2. Context Building

• ontext per file is built based on keywords and mode.
• Ensures AI understands which file/code block to reference.

3. LangChain + Google Gemini AI

• Different system prompts based on selected mode.
• Messages include previous AI responses for context-preserving chat.

4. Diff & Patch Engine

• AI suggestions are parsed for potential code changes.
• Suggested patches can be viewed or applied.

5. Security Scanning

• Static analysis-like checks for HIGH/MEDIUM/LOW severity issues.
• AI generates possible fixes with code diffs.

🎨 UI / UX Enhancements

Stateful Session: Context preserved for continuous interaction.

Color-Coded Analysis: Severity of issues shown with intuitive colors.

Loading Spinner: Shows progress during analysis.

Expandable Sections: AI insights and patches in expandable sections for better readability.

Best Practices & Notes

• Maximum file size per upload: 200MB.
• Context preserved per session, ensure to continue discussions in the same session.
• Google Gemini API quota limits apply — free tier allows limited requests daily.
• AI suggestions should be reviewed before applying in production.

🛠️ Tech Stack

Frontend / UI: Streamlit Backend / AI Client: LangChain, Google Gemini AI Language Parsing & Diffing: Python Environment Management: python-dotenv

📄 License

MIT License © [Shubham Sinha]

🔗 References

• LangChain Documentation

• Google Gemini AI

• Streamlit Documentation