Har_Report_Analyzer

namankabadi/Har_Report_Analyzer

3.2

If you are the rightful owner of Har_Report_Analyzer and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

Har_Report_Analyzer is a mini MCP server designed to analyze HAR files and provide detailed reports.

HAR Report Analyzer

This project provides a HAR (HTTP Archive) analyzer built on Model Context Protocol (MCP) standards. User Can Upload a HAR file as input and our MCP server will provide in depth analysis of HAR file with reports and deep analytical insights.

It consists of:

  • 🚀 MCP Backend Server (mcp_har_server.py) – FastAPI + Celery
  • 🎨 Streamlit UI (HAR_Analyser_UI.py) – interactive frontend
  • 📑 JSON/HTML reports with analysis & Plotly charts
  • 🔍 MCP Inspector integration for testing/debugging

✨ Features

  • Upload & analyze single or multiple .har files
  • Async task handling with Celery + Redis for large files
  • Rich metrics:
    • ✅ Top domains
    • ✅ Status & method breakdown
    • ✅ Largest & slowest requests
    • ✅ Requests over time
    • ✅ Performance & caching indicators
    • ✅ API request pass/fail
  • Auto-saves JSON reports in reports/
  • Visual charts using Plotly

🛠️ Setup

1. Clone this repo

2. Go to root directory using cd command in command promt

3. Install Necessary packages

pip install -r requirements.txt

Please create a new folder in the root directory with name .streamlit and create a secrets.toml file where you need to add Hugging Face Secrets like shown below and save the changes.

MCP_SERVER_URL = "http://127.0.0.1:8006" HF_Token = "YOUR SECRET TOKEN"

4. Start the MCP Server Backend using below command:

uvicorn mcp_har_server:app --host 127.0.0.1 --port 8006 --reload

Backend runs at → http://127.0.0.1:8006

📦 Endpoints

  • /health → Health check of MCP server
  • /upload_analyze → Upload a single HAR file
  • /upload_analyze_multi → Upload multiple HAR files
  • /task/{task_id} → Poll Celery async result
  • /reports/{id} → Get report JSON
  • /visual_report/{id} → Get Plotly visualization

5. Run HAR_Analyser_UI.py this file using below command:

streamlit run -m HAR_Analyser_UI.py

6. Upload the HAR File and get insights and report.

7: Test MCP Server using MCP Inspector :

Run this command to test MCP Server:

streamlit run MCP_Inspector.py

------------------------------