Leveling up OpenCode... and not in the way you would expect.
So I've been using OpenCode for a while now, and it's pretty cool. It's clean, minimal, effective, and not hacking other companies with rogue AI bots 😅. But there is one thing that I dislike about all of these AI tools besides people using them wrong: it's all 1 prompt, 1 agent at a time. Even with these new crazy models such as Kimi K3, Claude Fable 5, GPT Sol, DeepSeek V4 Pro, and the list…
OpenCode has been a handy tool for me, but its limitations are clear. One issue I've faced is the inability to handle multiple prompts or agents simultaneously. With the emergence of powerful AI models like Kimi K3, Claude Fable 5, GPT Sol, and DeepSeek V4 Pro, having reliable workflows or pipelines becomes crucial. While these advanced models have their own flaws, such as poor communication or outright lies, they highlight the need for better AI management.
Inspired by this, I decided to develop an open-source project, OpenFlow, to address these concerns. I discovered the concept of node-based workflows, which allowed me to fork OpenCode's harness and create a more efficient system. This new tool maintains the original BYOK (Bring Your Own Key) keys and supports any provider, but it introduces a novel twist: building a workflow with reusability in mind.
Users can create cards for agents, specifying their roles (planner, architect, coder, etc.), and connect them to other agents or chains.
This approach enables users to build complex or simple workflows as needed, allowing every agent to have a designated role and work together seamlessly. The flexibility of OpenFlow means it can be customized to suit individual needs. I continue to refine the project, and I encourage others to fork and adapt it to their own use cases. For more information, visit https://github.com/SeeRay11/OpenFlow.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
