Sample Open AI Swarm development team implementation with a log that shows communication between agents. Sample output.
Manager + Developer + Tester + Reporter
Team is reponsible to write and test python function based on instructions and function signature.
- Copy .env.sample to .env and fill in your Open AI key information
cp .env.sample .env
vi .env
- Install poetry packages
poetry install
poetry shell
python run.py <task>
task - subfolder in tasks folder that contains task details in format used by Aider Benchmark
It can be used to run exercism python exercises
- tasks
- <task>
| - .docs/instructions.md
| - <task>.py
| - <task>_test.py
To run with local Ollama model llama3.2 and qwen2.5-coder flip the ollama flag at the top of the run.py class