[mark_l_watson]

yes, whether for home experiments or at work, it is good practice (good hygiene) to be able to swap out both agentic harnesses and models. It is important to have a good strategy for exporting skills, etc.

[ig0r0]

yeah, I am using Pi right now, I switched from OpenCode

[carter2099]

I'm considering this right now. Is it very difficult to adapt to the new philosophy? Are you running mostly interactive or programmatic?

[ig0r0]

What do you mean by new philosophy? I use Pi the same way I used gpt or sonnet, just for simpler tasks.

[carter2099]

Pi is an agent, gpt and sonnet are LLMs. Opencode is an agent, and has a drastically different philosophy than Pi.

FWIW, I took the dive on Pi today and I’m really happy with my decision so far

[ig0r0]

Yeah, I meant using Pi with the local qwen model vs when I was using claude with sonnet or codex with gpt