What it may solve
DeepSeek Harness 推理强度增强Plugin:自动为 llm-pi-ai 下所有自定义模型补齐 x-opencode reasoning effort 档位(off→max 全七档,默认 high),写入 settings Composed of dsh 原生解析生效——新加的模型也会自动覆盖。
Imported third-party catalog description; not a Registry verification conclusion.