AI Agent Testing Framework – DataMind AI Assessment

Overview

Framework for testing an AI reservation agent using MCP tools.

• Agent: agent.py (stub for Claude AI agent)
• MCP Server: src/reservation_server.py (reservation tools)
• Test Cases: problems/001–004 (with validation scripts)
• System Prompt: prompts/system_prompt.txt (business rules)

---

Problems

Problem	Description
001	Minimal stub (passes example test)
002	Blocks January reservations
003	Adds services to reservations
004	Challenging multi-reservation + services scenario

---

Running Tests

From project root:

chmod +x docker-run.sh
./docker-run.sh problems/001
./docker-run.sh problems/002
./docker-run.sh problems/003
./docker-run.sh problems/004