Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

23 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

Tutorial Tester Agent

An AI-powered agent that automatically tests and validates tutorials using browser automation and intelligent decision-making.

πŸš€ Features

  • Browser Automation: Uses Playwright MCP Server for realistic browser interactions
  • AI-Powered Testing: LangGraph ReAct agent with Claude 4 Sonnet through SAP's Generative AI Hub
  • Tutorial Validation: Automatically executes tutorial steps and identifies issues

πŸ› οΈ Technology Stack

  • LLM: Claude 4 Sonnet access through SAP's Generative AI Hub (Cloud SDK for AI - Python)
  • Agent Framework: LangGraph with ReAct pattern
  • Browser Automation: Playwright MCP Server

πŸ“ Project Structure

tutorial-system-agent/
β”œβ”€β”€ tutorial-ai-agent/
β”‚   β”œβ”€β”€ agent/
β”‚   β”‚   └── tutorial_tester_agent.py    # Main agent
β”‚   └── data/
β”‚       β”œβ”€β”€ tutorials/                  # Tutorial markdown files
β”‚       └── output/                     # JSON results
β”œβ”€β”€ requirements.txt                    # Python dependencies
β”œβ”€β”€ docker-compose.yml                  # Container orchestration
└── Dockerfile                         # Container definition

🎯 How It Works

  1. Load Tutorial: Reads markdown tutorial files
  2. Initialize Agent: Creates LangGraph ReAct agent with MCP tools
  3. Execute Steps: Agent follows tutorial instructions in browser
  4. Validate Results: Checks for completion and identifies issues
  5. Generate Report: Saves structured JSON logs with timestamps

πŸ”§ Configuration

Key settings in tutorial_tester_agent.py:

  • TUTORIAL_FILE: Tutorial to test (default: "ailaunchpad-orchestration.md")
  • RECURSION_LIMIT: Max agent steps (default: 500)
  • MODEL_ID: LLM model (default: "anthropic--claude-4-sonnet")

About

An AI agent that automatically checks the tutorials from the SAP Tutorial system based on quality, relevance and functionality.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages