Routing OpenCode Tasks with Jev

说明与用途

该工具根据协调性、不确定性和后果等标准，自动判断是否使用强大的 AI Jev 模型或更轻量级模型，从而改进 OpenCode 的任务路由。

问题

Determine when to use a powerful AI model versus a lightweight one for routing tasks in OpenCode based on coordination, uncertainty, and consequence criteria.

结果

The tool enables dynamic routing of OpenCode tasks, selecting the appropriate AI model based on coordination, uncertainty, and consequence, improving efficiency and reducing manual oversight.

工作流

• 1. Define coordination, uncertainty, and consequence questions. 2. Create a tool in OpenCode that sends the approved plan to Jev via OpenRouter API. 3. Jev evaluates the plan against the questions and returns a decision. 4. Calculate complexity as the maximum of coordination and uncertainty scores. 5. Compare complexity to a consequence‑dependent threshold. 6. If complexity ≥ threshold, select the powerful model; otherwise, use the weaker model. 7. Return 'lite' or 'build' classification to the router.

证据

"Does successful implementation require reasoning about interactions or coordinated behavior across multiple parts of the system, rather than merely making localized changes in more than one file or module?"

来源

"If complexity is greater than or equal to the threshold value, the powerful model is selected; otherwise, the weaker one is selected."

来源