optimization

Per-round reasoning_effort optimizer for DeepSeek Harness (dsh): auto-downgrades tool-call reasoning for simple tool chains, lifting back for heavy work. Cuts thinking time between tool calls.

IsingQ toolkit for MCP and DeepSeek Harness: guided QUBO modeling, local validation, and user-confirmed solving. 面向 MCP 与原生 Agent 的 QUBO 建模与 IsingQ 求解工具集。