How tahaelkhaili/ia-agentic works
This repository is a collection of Agentic AI course labs and projects, not one unified runtime. Its central end-user workflow is the Evaluation_Final chat: a user submits a question, a LangGraph agent consults its LLM, optionally dispatches retrieval tools, and returns an answer. The README also documents separate learning branches for prompt engineering, agents, RAG, MCP, LangGraph, state, and human approval. Internal edges below follow excerpts or README statements; unsampled modules are shown as capabilities without inferred call paths.
Main entry points: nodes.py LLM and tool nodes · main.py Chat entry · graph.py Agent graph · llm.py Ollama model
Components 29
Final RAG Agent
- Chat entry
Evaluation_Final/main.py - Agent graph
Evaluation_Final/src/graph/graph.py - LLM and tool nodes
Evaluation_Final/src/graph/nodes.py - Conversation state
Evaluation_Final/src/graph/state.py - Ollama model
Evaluation_Final/src/model/llm.py - System prompt
Evaluation_Final/src/model/systemPrompt.py - Retrieval tools
Evaluation_Final/src/tools/tools.py - Document retriever
Evaluation_Final/src/RAG/retriever.py - Vector store
Evaluation_Final/src/RAG/vectorStore.py - Embedding model
Evaluation_Final/src/RAG/embeddings.py - Document loading
Evaluation_Final/src/RAG/loader.py - Text splitting
Evaluation_Final/src/RAG/splitter.py - Runtime settings
Evaluation_Final/config/config.py
Personal Chef Projects
- LangChain chef agent
Lab2-AgentsLangchain/TP_Agent_Personal_Chef.py - Chef chat app
TP-ChefPersonnel/main.py - Chef agent
TP-ChefPersonnel/src/Agent.py - Recipe RAG
TP-ChefPersonnel/src/RAG.py - Chef tools
TP-ChefPersonnel/src/Tools_Setup.py
Agent Patterns
- LangChain agents
Lab2-AgentsLangchain - Multi-agent Studio
Lab5-LangraphStudio - Context and state
Lab6-ContexteEtEtat - Human approval
Lab7-HumanInTheLoop - LangGraph agent
Lab9-AgentLangGraph/TP_Agent.py
AI Foundations
- Prompt engineering
Lab1-PromptIngineering - PDF and SQL RAG
Lab3-RAG - MCP integrations
Lab4-MCP
Graph Workflows
- LangGraph workflows
Lab8-WorkflowLangGraph - Agent HITL workflow
Lab9-AgentLangGraph/HITL_Workflow.py
Other components
- User
Connections 25
- User → Chat entry: asks question
Evaluation_Final/main.py - Chat entry → Agent graph: invokes graph
Evaluation_Final/main.py - Agent graph → LLM and tool nodes: routes execution
Evaluation_Final/src/graph/graph.py - LLM and tool nodes → Ollama model: calls model
Evaluation_Final/src/graph/nodes.py - LLM and tool nodes → System prompt: adds prompt
Evaluation_Final/src/graph/nodes.py - LLM and tool nodes → Retrieval tools: dispatches tools
Evaluation_Final/src/graph/nodes.py - Retrieval tools → Document retriever: searches documents
Evaluation_Final/src/tools/tools.py - Document retriever → Vector store: uses store
Evaluation_Final/src/RAG/retriever.py - Embedding model → Runtime settings: loads model setting
Evaluation_Final/src/RAG/embeddings.py - Ollama model → Runtime settings: loads model settings
Evaluation_Final/src/model/llm.py - User → Chef chat app: chats with chef
TP-ChefPersonnel/main.py - Chef chat app → Chef agent: invokes agent
TP-ChefPersonnel/main.py - Chef agent → Chef tools: uses tools
TP-ChefPersonnel/src/Agent.py - Chef tools → Recipe RAG: retrieves recipes
TP-ChefPersonnel/src/Tools_Setup.py - User → Prompt engineering: studies prompting
- User → LangChain agents: studies agents
- User → PDF and SQL RAG: studies RAG
- User → MCP integrations: studies MCP
- User → Multi-agent Studio: studies multi-agent graphs
- User → Context and state: studies state
- User → Human approval: studies approval
- User → LangGraph workflows: studies workflows
- User → LangGraph agent: studies LangGraph agent
- User → Agent HITL workflow: studies agent HITL
- User → LangChain chef agent: studies chef project