Dynamic Thinking Chain: A DSH Plugin That Automatically Adjusts Reasoning Depth Based on Task Difficulty
`dynamic-thinking-chain` is a DSH plugin that lets DeepSeek models automatically adjust their reasoning level based on task difficulty. It monitors task texts and instructions in `todo_write`, calculates a difficulty score, and dynamically maps it to one of four levels: `off/low/high/max`. `max` is used for complex planning, while `low` is used for simple tasks, saving tokens without compromising quality. The plugin has zero dependencies and supports custom threshold configuration, but be aware of the limitation of disabling DeepSeek’s thinking mode mid-session.
Read Moredsh-orquestrator: DeepSeek Harness Subagent Model and Inference Control Plugin
dsh-orquestrator is a plugin for DeepSeek Harness, designed to enforce subagent model selection, reasoning-effort caps, and output token limits to prevent cost overruns. Operating at the code level, it acts as a “gatekeeper” to ensure subagent configurations are applied. It provides native UI integration, allowing users to configure the subagent model, reasoning level, and orchestration skills before sending a message. The dsh process must be restarted after installation. This tool is suitable for developers who need fine-grained control over subagent resource consumption.
Read More