How ayushsharma-1234/ibm-agent-20 works
EduAI is a teacher-facing education assistant for Gujarat schools. Teachers use either the Streamlit interface or the CLI to request personalized learning plans, teaching materials, student assessment and analytics, or classroom support. The orchestrator dispatches requests to specialist agents; agents use shared prompt and result handling before generating responses through IBM WatsonX. The README documents the five principal agent areas; the tree also includes a test-paper agent. Source excerpts are partial, so links for that agent and some UI workflows are limited to documented structure.
Main entry points: orchestrator.py Orchestrator · base_agent.py Shared Agent Core · watsonx_client.py WatsonX Client · app.py Streamlit App
Components 15
User Interfaces
- Streamlit App
app.py - CLI
main.py
Request Routing
- Orchestrator
orchestrator.py
Education Agents
- Learning Recommendations
agents/learning_recommendation_agent.py - Lesson Planning
agents/lesson_plan_agent.py - Performance Analytics
agents/performance_analytics_agent.py - Assessment Feedback
agents/assessment_feedback_agent.py - Teacher Support
agents/teacher_assistant_agent.py - Test Paper
agents/test_paper_agent.py
AI Generation
- Shared Agent Core
agents/base_agent.py - WatsonX Client
watsonx_client.py - Model Configuration
config.py
Other components
- Teacher
- IBM IAM
- WatsonX Generation
Connections 21
- Teacher → Streamlit App: requests support
- Teacher → CLI: requests demo
- Streamlit App → Orchestrator: submits requests
- CLI → Orchestrator: submits requests
- Orchestrator → Learning Recommendations: dispatches
- Orchestrator → Lesson Planning: dispatches
- Orchestrator → Performance Analytics: dispatches
- Orchestrator → Assessment Feedback: dispatches
- Orchestrator → Teacher Support: dispatches
- Orchestrator → Test Paper: dispatches
- Learning Recommendations → Shared Agent Core: uses prompt helpers
- Lesson Planning → Shared Agent Core: uses prompt helpers
- Performance Analytics → Shared Agent Core: uses prompt helpers
- Assessment Feedback → Shared Agent Core: uses prompt helpers
- Teacher Support → Shared Agent Core: uses prompt helpers
- Test Paper → Shared Agent Core: uses prompt helpers
- Shared Agent Core → WatsonX Client: requests generation
- Orchestrator → WatsonX Client: classifies intent
- WatsonX Client → Model Configuration: reads settings
- WatsonX Client → IBM IAM: gets token
- WatsonX Client → WatsonX Generation: sends generation request