Multi-role parallel argumentation for DeepSeek Harness: runs the real Codex and Claude CLIs in parallel, has the DSH model (Judge) write a third-party verdict back into the conversation, plus Obsidian-style single-agent direct chat and per-role model config.
Install
# from npm (prebuilt)
dsh plugin --profile web add dsh-multi-role-debate
# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)
dsh plugin --profile web add github:jiang12345-code/dsh-multi-role-debate#path:/dsh-multi-role-debate
Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time — pnpm blocks those until you allow them, so an install can stop with ERR_PNPM_GIT_DEP_PREPARE_NOT_ALLOWED or ERR_PNPM_IGNORED_BUILDS; dsh prints the exact key to add under allowBuilds in your profile’s pnpm-workspace.yaml, and the install works on the next run. Allowing a build is a trust decision: only install sources you trust, and pin a commit (github:owner/repo#sha).
README
This plugin publishes its README in Chinese only.
这是多角色并行论证 + 单 Agent 直接对话 + DSH Judge 汇总 + 结果回对话的 DSH 聚合插件包。它是一个"一次性安装"入口：本身只提供 cordis.patch.yml（注册 3 个组件 bundle），并依赖下面 3 个组件包：
| 组件包 | 作用 |
|---|---|
dsh-codex-agent |
host 实体：长驻真实 codex app-server，提供 codexAgent 服务 |
dsh-claude-agent |
host 实体：长驻真实 claude-agent-sdk，提供 claudeAgent 服务 |
multi-role-debate |
编排 + 前端：多角色论证 / 对话流触发 / DSH 汇总 / 结果回对话 / 直接对话 UI |
安装本聚合包，上面 3 个组件会一起注册、一起生效（见仓库根 README.md 的安装步骤）。
前置依赖
- DSH（DeepSeek Harness）web profile。
- 本机已安装 Codex CLI（
@openai/codex）与 Claude Agent SDK（@anthropic-ai/claude-agent-sdk）——本插件是围绕这两个真实 CLI 的薄封装，不内置二进制。
Links
More in this category
Q00/ouroboros#integrations/dsh-plugin★ 5825
Config-only bundle that mounts Ouroboros through the DSH MCP client, exposing 36 interview, Seed, execution, evaluation, and evolution workflow tools in DSH.
huangruiteng/loopx#dsh-loopx-plugin★ 5818
LoopX, a provider-neutral, local-first state kernel and control plane for long-horizon agents: keeps Goal, Todo, gate, evidence, quota, recovery, and handoff state above DeepSeek Harness, while the plugin bootstraps the CLI and skills, admits bounded same-session continuation, and adds a loopback GoalBar for the exact bound loop.
chuspeeism/dashi-taskboard#deepseek-harness★ 3070
Embeds the active installed Codex Taskboard runtime in the DeepSeek Harness sidebar, using its launcher runtime descriptor instead of a fixed port.
NanmiCoder/dsh-agent-teams★ 1596
AgentTeams multi-agent teams.
tong-io/tongflow#dsh-tongflow★ 1020
TongFlow film-crew studio for image, voice, music and video production: the agent writes per-asset TongFlow workflow files (.tongflow.json) that run through TongFlow plugins, with an embedded workflow canvas, a shot/character/take project layout and a manga-drama template; sessions starting with @tongflow open the Studio view.
EthanYoQ/AI-Novel-Writer#dsh-ai-novel-writer★ 824
Installs a dedicated AI novel-writing preset and workbench: revisioned local project assets, a compact side drawer, and native approval-gated single-file changes.
Community comments
Comments are public GitHub Discussions. Loading them connects to GitHub and Giscus; a GitHub account is required to post.