maker-framework

SKILL.md

Transform unreliable single-model inference into robust, verifiable reasoning through maximal decomposition, parallel consensus voting, and systematic error filtering.

Decompose complex tasks into atomic, independently-executable subtasks forming a DAG.

Execute each subtask with m parallel agents; accept when one result leads by k votes.

使用 MAKER（用于推理的最大代理知识引擎）编排可靠的多代理推理。实施三支柱架构，将概率性法学硕士输出转换为确定性、可验证的结果。当任务需要高可靠性、并行共识投票或系统错误检测时使用。触发器包括可靠性关键任务、多步推理链、基于共识的验证、并行代理执行或显式 MAKER 调用。 来源：zpankz/mcp-skillset。

查看原文