Skip to content
Plugin

dsh-routing-suite

yjh0511087,017196Updated 1mo ago

Open on GitHub

dsh-routing-suite — injector + router-standard kit: install the runtime injector first, then the task-aware reasoning-mode router preset (measured P1-P23).

Practical score

80 / 100

Setup

Ready to use

License

MIT

Language

JavaScript

Install

Quick start

Copy the command, or hand the whole task to your harness agent.

dsh 不在 PATH 时:npx '@deepseek-ai/dsh' plugin --profile web add .\injector
Install with your agent

Please install the DeepSeek Harness plugin "dsh-routing-suite" for me. - GitHub repository: yjh051108/dsh-routing-suite (https://github.com/yjh051108/dsh-routing-suite) - Type: a Cordis plugin (installed into a DSH profile) - It should work out of the box with…

From the README

A cleaned excerpt from the project's own documentation.

dsh-routing-suite — 注入器 × 思维模式路由 × 分级模式 套装 一个仓库装齐三件套: 运行时注入器 (免重启运行时管理层)+ 思维模式路由预设 (任务感知推理模式,P1-P23 实测)+ 分级任务协议 (脑暴出题 → 规格化计划 → 打卡制 → 组收官 → 终验,含红队门与审计端点)。 中文 English 组件入口(快速跳转) 组件 说明 入口 --- --- --- injector 运行时注入器: dev 工具全家桶(注入/热重载/卸载/侧挂转正/路由自愈) injector/README.md · 文档 preset 思维模式路由预设(router-standard / router-spec,按模型 persona 路由) preset/README.md · 实验报告 graded 分级模式:会话级两级任务协议(6 工具+三模式+小类模式粒度+红队门+审计端点+超级面板) graded/README.md · 架构 · 理论 · 实测数据 研究与数据(graded) "模型也会偷懒——我们测到了,还做了对抗" —— 分级模式插件的公开研究位: - 🔬 模型注意力懈怠:观测笔记——长程任务里验证勤勉度衰减的真实分段数据(无协议链晚段"1 工具/0 读图" vs 协议链全程在线)+对抗机制对照 - 📊 协议 vs 无协议:实测对比——同型 3D 任务三会话对比表+ASCI I 图 - 🧪 测量工具—— node scripts/measure.mjs 你的会话.jsonl :任何人可对自己会话复算上述指标(脱敏:…

Open on GitHub

Project details

by
yjh051108
License
MIT
stars
7,017
Forks
196
Updated
Sep 18, 2026
Created
Aug 14, 2026
Language
JavaScript
Indexed from
topic

Related entries

Similar tools from the same corner of the ecosystem.

View all
Plugin

dsh-toolbox-ui

Imzl-zl

Toolbox panel for DeepSeek Harness Web: MCP manager (tools & JSON Schema, market, import, workspace copy) and skills manager (scopes & shadowing, enable/disable, model auto-invocation, skillhub.cn market).

agent-skillscordisdeepseek-harness
46Ready to use
82/100
Plugin

dsh-expert-team

yangdcm

Role-based multi-agent expert team for DeepSeek Harness: one sentence in, a staged and gated team delivery out. 12 role subagents, 9 gated phases, shared-workspace artifacts, zero runtime dependencies.

agent-orchestrationai-agentscode-generation
18Ready to use
82/100
Plugin

dsh-preset-workshop

Zai-G

预设工坊 · Preset Workshop:为 DeepSeek Harness 管理人设、系统提示词、工具、MCP与技能,并预览请求组成。

deepseek-harnessdshdsh-plugin
15Ready to use
82/100