Submit a project
dsh-plugin-jev-effort-selector
By justhalfbit
Source description: DeepSeek Harness (DSH) 推理等级自动选择插件：由 Jev System One 模型判断每条消息值多少思考量，按模型声明的等级自动推导档位，上下文信封让「继续」这类追问继承话题深度，低置信度向上取，任何失败都静默沿用原等级。 | Jev-driven reasoning effort per message: per-model ladders derived from what each model advertises, a fixed-size c
Repository description
The repository describes its purpose as: DeepSeek Harness (DSH) 推理等级自动选择插件：由 Jev System One 模型判断每条消息值多少思考量，按模型声明的等级自动推导档位，上下文信封让「继续」这类追问继承话题深度，低置信度向上取，任何失败都静默沿用原等级。 | Jev-driven reasoning effort per message: per-model ladders derived from what each model advertises, a fixed-size c
Source evidence
The daily check found an explicit Jev / TypeSafe AI connection in the original README and implementation files at the same commit.
让 [Jev](https://typesafe.ai) System One 模型替你决定每条消息该用多深的推理。
Automatically source-checked: public repository, pinned README, and implementation files. Not run, benchmarked or security-audited.
