Skip to content

README.md

Latest commit

 

History

History
30 lines (21 loc) · 1.45 KB

README.md

File metadata and controls

30 lines (21 loc) · 1.45 KB

EdgeInsight: AI-Powered Repository Auditor

EdgeInsight is a full-stack AI application built entirely on the Cloudflare Developer Platform. It performs architectural and security audits on GitHub repositories in seconds.

🔗 Live Demo: [edgeinsight.pages.dev]

⚡ Tech Stack (Cloudflare Native)

This project strictly adheres to the Cloudflare "North Star" architecture:

  • Frontend: Next.js (App Router) deployed on Cloudflare Pages.
  • AI Inference: Workers AI running Llama 3.1 8B Instruct (optimized for edge latency).
  • Orchestration: Cloudflare Workflows to handle long-running fetch & analysis tasks.
  • State Management: Durable Objects for real-time status updates and result persistence.
  • Backend: Cloudflare Workers (Hono.js) as the API gateway.

🚀 How It Works

  1. User Input: User submits a GitHub URL via the Next.js frontend.
  2. Workflow Trigger: The Worker initiates an asynchronous AuditWorkflow.
  3. Step 1 (Fetch): The Workflow fetches the raw code from GitHub (in-memory to avoid storage limits).
  4. Step 2 (Analyze): Llama 3 analyzes the codebase for security risks, architecture, and tech stack.
  5. Step 3 (Persist): Results are saved to a Durable Object.
  6. Realtime UI: The frontend polls the Durable Object to show a live "Processing" state and the final report.

🛠️ Local Development

  1. npm install
  2. npx wrangler dev (Backend)
  3. npm run dev (Frontend)