How I write code with agents using OpenCode 2/13/26
Автор: Scout
Загружено: 2026-02-14
Просмотров: 183
Описание:
This is a short 20-minute video that I recorded for a friend showing how I code a feature using OpenCode.
OpenCode is a local agentic CLI desktop application that allows you to essentially pair program with an agent to write code. One of its strengths is the ability to switch between models and not be dependent on any one model. It also has many other features like running sub-agents and slash commands and hooks or plugins and just so much more.
I use OpenCode in a vanilla mode because I want to get real signal between the models to determine which model does certain things, good, bad, or it doesn't matter, and how fast they are. The main thing I want to get across in this video, it is the fact that the pattern, regardless of agentic harness, is valid. By researching and planning first, then clearing the context and executing on the plan with sub-agents in parallel, then finally clearing the context again and doing a full code review, results in a solid flow. I just wanted to share that with some fellow team members.
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
