Skip to content

QUICK_REFERENCE.md

Latest commit

 

History

History
193 lines (149 loc) · 4.05 KB

QUICK_REFERENCE.md

File metadata and controls

193 lines (149 loc) · 4.05 KB
Raw

🚀 Quick Reference Card

Start All Services (Copy & Paste)

Windows PowerShell

Terminal 1 - Planner Agent:

cd Planner-Agent\agent\planner
.\venv1\Scripts\Activate.ps1
uvicorn main:app --reload --port 8001

Terminal 2 - Dataset Agent:

cd Dataset_Agent\agents\dataset
.\venv\Scripts\Activate.ps1
python main.py

Terminal 3 - MCP Server:

cd mcp_server
.\venv\Scripts\Activate.ps1
uvicorn main:app --reload --port 8000

Terminal 4 - Backend:

cd backend
npm start

Terminal 5 - Frontend:

cd frontend
npm run dev

Quick Health Checks

# Planner Agent
curl http://127.0.0.1:8001/health

# MCP Server
curl http://127.0.0.1:8000/health

# Frontend
# Open: http://localhost:5173

Service Ports

Service Port URL
Planner Agent 8001 http://127.0.0.1:8001
Dataset Agent 8002 http://127.0.0.1:8002
MCP Server 8000 http://127.0.0.1:8000
Backend 4000 http://localhost:4000
Frontend 5173 http://localhost:5173

Common Commands

Fix Planner Agent Package Issue

cd Planner-Agent/agent/planner
pip uninstall -y supabase httpx gotrue
pip install -r requirements.txt

Check Database

-- Recent projects
SELECT * FROM projects ORDER BY created_at DESC LIMIT 5;

-- Recent logs
SELECT * FROM agent_logs ORDER BY created_at DESC LIMIT 10;

Test Planner Agent Directly

curl -X POST http://127.0.0.1:8001/agents/planner/handle_message \
  -H "Content-Type: application/json" \
  -d "{\"user_id\":\"test\",\"session_id\":\"test\",\"message_text\":\"Train a cat classifier\"}"

Test Messages

Try these in the ML Chat:

  1. "Train a model to classify plant diseases"
  2. "Create an image classifier for skin cancer detection"
  3. "Build a model to identify different types of flowers"
  4. "Train a cat vs dog classifier"

Troubleshooting Quick Fixes

Error Fix
Planner Agent won't start pip install supabase==2.9.0 httpx==0.27.0
MCP Server validation error Already fixed in code
Service unavailable Check if Planner Agent is running
Gemini API error Check GEMINI_API_KEY in .env files
Projects not showing Refresh page or switch tabs

File Locations

Auth_System_MCP-main/
├── Planner-Agent/agent/planner/
│   ├── main.py              # Planner service
│   ├── .env                 # Planner config
│   └── requirements.txt     # Planner packages
│
├── mcp_server/
│   ├── main.py              # MCP orchestrator
│   └── .env                 # MCP config
│
├── backend/
│   ├── src/routes/ml.js     # ML routes
│   └── .env                 # Backend config
│
└── frontend/
    ├── src/components/
    │   └── MLChatBot.jsx    # ML chat interface
    └── .env                 # Frontend config

Environment Variables Checklist

Planner Agent (.env)

  • SUPABASE_URL
  • SUPABASE_KEY
  • GEMINI_API_KEY
  • LOG_LEVEL

MCP Server (.env)

  • SUPABASE_URL
  • SUPABASE_SERVICE_ROLE_KEY
  • GEMINI_API_KEY
  • PLANNER_AGENT_URL

Backend (.env)

  • PORT
  • MCP_SERVER_URL
  • FIREBASE_PROJECT_ID
  • GOOGLE_APPLICATION_CREDENTIALS

Success Indicators

✅ Working When:

  • All 4 services start without errors
  • Health checks return 200 OK
  • Can create projects via ML Chat
  • Projects appear in UI
  • Agent logs visible in database
  • Status shows "Pending Dataset"
  • Progress bar at 25%

Documentation Index

Document Purpose
TESTING_GUIDE.md Complete testing instructions
PLANNER_AGENT_INTEGRATION.md Integration details
INTEGRATION_COMPLETE.md Success summary
SETUP_GUIDE.md Initial setup
FEATURE_STATUS.md Feature completion status
QUICK_START.md Quick start guide
QUICK_REFERENCE.md This file

Need Help? Check TESTING_GUIDE.md for detailed troubleshooting!