UNPKG

research-informed-ai-workflow

Version:

Research-Informed AI Workflow Module for comprehensive codebase analysis and improvement using Gemini AI and GitHub Copilot

122 lines (106 loc) 2.88 kB
# Example: Node.js Web Application Analysis This example demonstrates how to use the Research-Informed AI Workflow Module with a Node.js web application. ## Project Structure ``` example-nodejs-app/ ├── src/ │ ├── controllers/ │ ├── models/ │ ├── routes/ │ └── app.js ├── tests/ ├── package.json ├── README.md └── .gitignore ``` ## Custom Configuration Files ### bundles.example.nodejs.json ```json { "tasks": [ { "id": "comprehensive-analysis", "name": "Comprehensive Analysis", "instructionsFile": "tasks/01_comprehensive_analysis.md", "bundle": [ "src/**/*.js", "package.json", "README.md", "tests/**/*.js" ] }, { "id": "security-vulnerability-analysis", "name": "Security Vulnerability Analysis", "instructionsFile": "tasks/02_security_vulnerability_analysis.md", "bundle": [ "src/**/*.js", "package.json", "tests/**/*.js" ] }, { "id": "testing-system-analysis", "name": "Testing System Analysis", "instructionsFile": "tasks/03_testing_system_analysis.md", "bundle": [ "tests/**/*.js", "package.json" ] } ] } ``` ### repomix.config.example.nodejs.json ```json { "output": { "filePath": "analysis-bundle.txt", "style": "plain", "removeComments": false, "removeEmptyLines": false }, "include": [ "**/*.js", "**/*.json", "**/*.md", "!node_modules/**", "!dist/**", "!build/**" ], "ignore": { "useGitignore": true, "useDefaultPatterns": true, "customPatterns": [ "*.log", ".env*", "coverage/**" ] } } ``` ## Usage Instructions 1. **Copy the configuration files** to your project root: ```powershell Copy-Item "examples\bundles.example.nodejs.json" ".\bundles.json" Copy-Item "examples\repomix.config.example.nodejs.json" ".\repomix.config.json" ``` 2. **Run the setup script**: ```powershell .\setup.ps1 ``` 3. **Generate analysis bundles** for specific tasks: ```powershell .\Generate.ps1 -TaskID "comprehensive-analysis" ``` 4. **Use with Gemini AI**: - Upload the generated `analysis-bundle.txt` to Gemini - Request analysis using the corresponding task instructions - Apply the insights using GitHub Copilot ## Expected Output After running the generation script, you'll get: - `analysis-bundle.txt` - Complete codebase bundle for AI analysis - Task-specific instructions in the `tasks/` directory - Ready-to-use analysis workflow ## Customization Modify the bundle patterns in `bundles.json` to match your specific project structure and analysis needs.