workshop-mcp-server by gangwgr - MCP Server

Workshop MCP Server

AI-Powered Development Assistant with code review, GitHub PR automation, OpenShift test generation & execution, must-gather analysis, and cluster debugging capabilities.

Features

Code Review - Line-by-line analysis with security, performance, and quality checks
PR Review - Automated GitHub PR reviews with comments
OpenShift Testing - Test generation (Gherkin/YAML/Go/Shell) and execution
Must-Gather Analyzer - Cluster health assessment with AI-powered diagnostics
Cluster Debugger - Intelligent issue analysis with fix recommendations
Web GUI - Beautiful interface for all tools

Quick Start

🚀 Automated Setup (Recommended)

# 1. Clone repository
git clone https://github.com/gangwgr/workshop-mcp-server.git
cd workshop-mcp-server

# 2. Auto-configure for your AI system
./setup-ai.sh all        # Configure all AI systems (Gemini, OpenAI, Claude)
# OR choose specific:
# ./setup-ai.sh gemini   # Only Google Gemini
# ./setup-ai.sh openai   # Only OpenAI GPT-4
# ./setup-ai.sh claude   # Only Claude Desktop

# 3. Start web GUI
./setup-ai.sh start

# 4. Open browser: http://127.0.0.1:8080

📝 Manual Setup

# 1. Clone and setup
git clone https://github.com/gangwgr/workshop-mcp-server.git
cd workshop-mcp-server
python3 -m venv .venv
source .venv/bin/activate

# 2. Install dependencies
pip install -e .
pip install -r integrations/requirements-integrations.txt

# 3. Configure API keys (create .env file)
cp .env.template .env
# Edit .env and add your API keys

# 4. (Optional) Authenticate for GitHub/OpenShift features
gh auth login
oc login https://api.your-cluster.example.com:6443

# 5. Run web GUI
cd web_gui
python app.py

# 6. Open browser: http://127.0.0.1:8080

For detailed setup, see | Prerequisites:

Usage

Web GUI (Recommended)

Access at http://127.0.0.1:8080 for an intuitive interface to all tools.

MCP Server (Claude Desktop)

pip install -e .
python -m workshop_mcp_server.server

Configure Claude Desktop to use this server (see ).

AI System Integrations

Supports multiple AI systems out of the box:

🤖 Google Gemini - Direct API & Function Calling
🧠 OpenAI GPT-4 - Function Calling & Interactive Chat
💬 Anthropic Claude - MCP Protocol (Claude Desktop)
🔗 LangChain - Universal integration for ANY LLM

Auto-configuration available:

./setup-ai.sh all  # Configure all AI systems with guided setup

# Gemini Example
from integrations.gemini_integration import GeminiMCPClient
client = GeminiMCPClient()
result = client.review_code(code="def test(): pass", with_gemini_analysis=True)

# GPT-4 Example
from integrations.openai_integration import OpenAIMCPClient
client = OpenAIMCPClient()
response = client.chat("Debug my OpenShift API server")

# LangChain (Universal - works with ANY LLM)
from integrations.langchain_integration import create_mcp_agent
from langchain_google_genai import ChatGoogleGenerativeAI
agent = create_mcp_agent(ChatGoogleGenerativeAI(model="gemini-pro"))
response = agent.run("Review this code for security issues: ...")

See for complete guide.

Available Tools

Tool	Description
`review_code_line_by_line`	Line-by-line code review with severity filtering
`post_pr_review_comments`	Post GitHub PR review comments
`generate_ocp_test_case`	Generate OpenShift tests (Gherkin/YAML/Go/Shell)
`execute_ocp_test_step_by_step`	Execute tests with real-time progress
`debug_ocp_test_failure`	Intelligent test failure analysis
`analyze_mustgather_bundle`	Analyze must-gather bundles with AI
`debug_openshift_cluster`	Cluster debugging with diagnostics

Architecture

Users (Web Browser / Claude Desktop)
           ↓
Web GUI / MCP Server / CLI Tools
           ↓
Tool Layer (Python Functions)
           ↓
GitHub API / OpenShift Cluster / Must-Gather Files

Tech Stack: Python 3.8+, Flask, MCP Protocol, GitHub CLI, OpenShift CLI

Deployment

Production (Gunicorn)

pip install gunicorn
cd web_gui
gunicorn -w 4 -b 0.0.0.0:8080 app:app

Docker

docker build -t workshop-mcp-server .
docker run -p 8080:8080 -v ~/.kube/config:/root/.kube/config workshop-mcp-server

OpenShift

oc new-app python:3.9~https://github.com/your-username/workshop-mcp-server.git \
  --context-dir=web_gui --name=mcp-server
oc expose svc/mcp-server

Troubleshooting

Issue	Solution
Port 8080 in use	`lsof -i :8080` then `kill -9 <PID>`
Module not found	Activate venv: `source venv/bin/activate`
GitHub auth failed	`gh auth logout && gh auth login`
OC cluster failed	`oc login https://api.cluster.com:6443`

See for more help.

Contributing

Fork the repository
Create feature branch
Make changes with tests
Submit pull request

License

MIT License - see file.

Support

Documentation: | |
Issues: https://github.com/your-username/workshop-mcp-server/issues

Built for DevOps, SRE, and Developer Teams