DeepSeek Harness plugin

dsh-delegation-workflow

主规划→子执行→主审查 的委派工作流插件,支持为子代理动态指定模型与思考强度

Jump to install

Source facts

Repository
Wh1stle05/dsh-delegation-workflow
Latest update
Aug 16, 2026
Category
Workflow & Automation
GitHub stars
0
Format
plugin
Catalog evidence
Upstream dsh.bundle evidence
Evidence path
package.json#dsh.bundle
Checked against
0.1.0-rc.8
Upstream check date
2026-08-20

This evidence comes from the upstream catalog. This site has not installed, run, or security-reviewed the plugin.

Install

Start with a prompt that asks an agent to review the GitHub repository and source. Switch to the command if you want to install it yourself.

Copy this prompt into DSH, Codex, or another agent and ask it to review the GitHub repository and source first.

Do not install or run any commands yet. Read this plugin's GitHub repository, README, and relevant source code. Then answer the questions below clearly and directly so I can decide whether it fits my needs:

1. What is this plugin, and what problem does it solve?
2. Who is it for, and what are its typical use cases?
3. How is it used after installation? Include one minimal example.
4. What known limitations or privacy, security, compatibility, or maintenance risks does it have?
5. Give a clear recommendation: recommend, conditionally recommend, or do not recommend, with reasons.

Distinguish statements documented by the repository, inferences from source code, and unknowns. If evidence is insufficient, say so explicitly. Do not guess or simply repeat the README.

GitHub: https://github.com/Wh1stle05/dsh-delegation-workflow
Plugin: dsh-delegation-workflow
Author: Wh1stle05

Check the source files

Read the README and other files from this plugin directory before installing.

File explorer3 files
README.mdSource · read only

dsh-delegation-workflow

协作式委派工作流插件:主规划 → 子执行 → 主审查。为 DeepSeek Harness 提供一个 delegate_with_model 工具,让主 agent 把计划中的一个步骤委派给子代理执行,并显式指定该子代理的运行模型与思考强度

功能

  • 新增工具 delegate_with_model,参数:

- description — 委派任务的简短描述(3-5 词,用于展示)。 - prompt — 自包含的子代理任务说明(子代理不共享父代理上下文)。 - provider — 子代理的 LLM 模型后端,缺省回落 settings 默认值。 - model — 子代理使用的模型 id,缺省回落 settings 默认值。 - reasoning_effort — 子代理的推理强度,动态校验目标模型真实支持的档位。 - run_in_background — 是否以后台任务运行(返回 job id,可用 job_output/job_kill 管理)。

  • Web 端新增 /delegate-model 斜杠命令:先选模型、再选思考强度,写入 $DSH_HOME/settings.yamldelegation-workflow: 段。
  • 「通用设置」新增「子代理默认模型」条目,模型与思考强度分成两个选择器,模型条目显示提供商信息。
  • 与框架内置 subagent 工具的区别:在委派边界上把调用方指定的 provider / model 写入子代理 AgentOptions,并在子代理发出首个请求前通过 installModelSelection 注入 reasoning_effort

安装

前提:已安装 DeepSeek Harness(dsh 命令)。dsh plugin 命令底层转发给 pnpm,因此可直接从 GitHub 安装本仓库:

# 安装最新版
dsh plugin --profile web add github:Wh1stle05/dsh-delegation-workflow

# 或固定到某个版本 tag
dsh plugin --profile web add github:Wh1stle05/dsh-delegation-workflow#v0.2.0

然后把本仓库 cordis.patch.yml 里的 insert 片段并入你的 profile cordis.patch.yml(见下方「配置」),并重启 dsh 服务使插件源码生效(插件 src/index.js 的改动需进程重启才会加载,cordis.patch.yml 的 config 可热重载)。

配置

cordis.patch.yml 中的插件条目:

- insert:
    - id: delegation-workflow
      name: dsh-delegation-workflow
      config:
        subagentProvider: spawn   # 子代理执行后端(spawn / fork)
        provider: opencode-go     # 子代理的 LLM 模型后端
        model: deepseek-v4-flash
        reasoningEffort: high
        toolName: delegate_with_model

> provider / model / reasoningEffort 仅作为 settings namespace > delegation-workflow 的 composition base;可在 Web 端通过 > /delegate-model 斜杠命令或「通用设置 → 子代理默认模型」修改, > 修改会持久化到 $DSH_HOME/settings.yaml

> 注意:subagentProviderprovider 是两个不同概念: > - subagentProvider执行后端(transport),由 dsh-subagent-{spawn,fork}-in-process 注册,内置值为 spawn / fork; > - provider LLM 模型后端(如 opencode-go / deepseek-official),通过 agentOptions.provider 注入。 > > 二者不可混用,否则会报 no subagent provider registered for ...

使用

主 agent 规划任务后,用 delegate_with_model 委派子代理执行,子代理返回后由主 agent 审查:

主 agent: todo_write 拆解计划
主 agent: delegate_with_model(步骤 1, model=..., reasoning_effort=..., run_in_background=false)
子代理:   执行并返回结果
主 agent: 对照计划审查,直到验收通过

详见 AGENTS.md 中的工作流规则。

许可

[MIT](./LICENSE)