News for hackers. Stuff that matters.

piscaries

Author here. The original article is here: https://piscaries.github.io/posts/codex-and-pi-build-the-sam...

I had two goals: 1. try out Orca for running a team of coding agents 2. See how Pi, a minimal open-source agent, generates codes against a popular one like Codex.

So I ran the same small team (spec, design, build, with a shared reviewer) twice in Orca. Once on Codex with GPT-5.6. The other on Pi with GLM-5.3. Both built chess coach apps. Then I let their chess coaches compete. Codex finished faster, but in a 20-game match Pi's engine won 18 and drew the other 2.

Prompts, both apps, tests, judging and match logs are in the repo: https://github.com/piscaries/codex-vs-pi-in-orca

Happy to answer questions about the setup and welcome feedback.

Agree, you can't spec what you don't know yet. Two things could help: Let the agent grill you before writing the spec. Answering its questions surfaces a lot of the unknowns. Then have a different model grill the plan, with questions specific to your product. In one case of mine, that caught a design issue which the generic review missed.