OpenViking
volcengine
Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
PROJECT TOPICS
INSTALL REFERENCE
dsh plugin --profile web add github:Asher-2000/dsh-expert-mode
该命令指向仓库当前默认分支;尚无绑定当前 commit 的完整验证结果。
PROJECT README
An agent preset for DeepSeek Harness (DSH): a "chief coordinator" plus 11 domain-expert subagents with automatic task delegation.
中文版见 README.zh.md
"Expert Mode" is a DSH agent preset. Once mounted, your agent becomes a chief coordinator commanding 11 domain-expert subagents. When a task arrives, the coordinator decides which domain it belongs to, delegates it to the best expert, and consolidates the final delivery.

Select "Expert Mode" preset in the DSH workspace

Tasks are automatically delegated to multiple domain-expert subagents in parallel, with real-time token usage and duration
Who is it for
| Expert | Delegation Tool | Domain |
|---|---|---|
| Data Analyst | expert_data_analyst |
data processing, statistics, visualization |
| Copywriter | expert_copywriter |
marketing copy, content creation |
| Legal Review | expert_legal_review |
contracts, legal risk |
| Product Manager | expert_product_manager |
requirements, product planning |
| Frontend Dev | expert_frontend_dev |
web frontend implementation |
| UI/UX Design | expert_uiux_design |
interface design, interaction |
| Architect | expert_architect |
system design, tech selection |
| Social Media Ops | expert_social_media |
multi-platform content distribution |
| Growth Hacker | expert_growth |
growth strategy, conversion |
| Quant Finance | expert_quant_finance |
quant models, financial analysis |
| Finance | expert_finance |
financial analysis, reporting |
Option 0: dsh plugin add (recommended)
dsh plugin --profile web add github:Asher-2000/dsh-expert-mode
Option 1: git clone
mkdir -p ~/.dsh/.agent-presets
git clone https://github.com/Asher-2000/dsh-expert-mode.git ~/.dsh/.agent-presets/expert-mode
Want the English version? The repo ships an expert-mode-en/ preset (Chief Coordinator + 11 experts, fully English). After cloning, copy it into your presets directory:
cp -r ~/.dsh/.agent-presets/expert-mode/expert-mode-en ~/.dsh/.agent-presets/expert-mode-en
Then pick "Expert Mode" (English) or 「专家模式」 (Chinese) in the DSH preset picker.
Option 2: manual download
Download the latest release from Releases, then place preset.yml + agent.cordis.yml into ~/.dsh/.agent-presets/expert-mode/.
After installation, select the "Expert Mode" preset when creating a new session in the DSH Web GUI; or specify it from the command line:
dsh --profile headless "Analyze this data and give recommendations" --preset expert-mode
Note: the preset selector lives in the Agent preset dropdown of the session creation screen. Exact flags depend on your installed DSH version.
When delegating, state the goal, inputs, and expected output clearly so the expert subagent works independently; when consolidating, keep the expert's conclusions and data evidence intact without rewriting.
.
├── preset.yml # preset metadata (name + description)
├── agent.cordis.yml # cordis composition: coordinator persona + 11 expert subagent tools
├── README.md
├── README.zh.md
└── LICENSE
Q: Does Expert Mode consume extra model quota? A: Delegating to an expert subagent does incur the subagent's model calls (standard DSH subagent mechanics); simple tasks answered directly by the coordinator cost nothing extra.
Q: Can I add my own experts?
A: Yes. Copy any expert entry from agent.cordis.yml, adjust the tool name and persona.
Q: Relationship to the official standard preset? A: Built on top of the official standard preset composition, keeping the full toolset and adding an expert-delegation layer.
dsh deepseek-harness agent-preset expert-mode multi-agent subagent ai-agent dsh-plugin
MIT License — see LICENSE.
CLASSIFICATION EVIDENCE
系统优先读取 GitHub Topics,再与站内分类词典和词根规则比对。当前命中: multi-agent、subagent。