Introduction
For your final capstone, you will build a complete Multi-Agent System using crewAI. You will architect an autonomous "Software Engineering Team" that can write, review, and execute code based on a simple user prompt.
Requirements
- Define Agents: Create at least three distinct agents: A Senior Python Developer (writes the code), a Quality Assurance Engineer (reviews and tests the code), and a Product Manager (oversees the goal and writes the final documentation).
- Equip Tools: Give the Developer agent a tool to execute Python code safely (or write it to the filesystem) and read the output/errors.
- Design the Workflow: Set up a sequential task pipeline: Manager defines specs -> Developer writes and runs code -> QA reviews and sends back errors if any -> Manager outputs the final project summary.
- Execution: Provide the crew with a complex task, like "Build a Python script that fetches data from a public weather API, parses the JSON, and creates an ASCII art chart of the temperature."
Hints
- Use LangChain's built-in File System tools or create custom tools for the agents to read and write files.
- Instruct the QA agent specifically to check for edge cases and handle exceptions.