dsh-meta-orchestrator
Run the following command in DeepSeek Harness:
dsh plugin install jiruidai/dsh-meta-orchestrator
Paste the following prompt into your AI chat to install this plugin:
In DeepSeek Harness, install the plugin by running `dsh plugin install jiruidai/dsh-meta-orchestrator` (source: https://github.com/jiruidai/dsh-meta-orchestrator) into the web profile, then restart dsh.
About this plugin
dsh-meta-orchestrator solves the critical pain point of rigid, inflexible workflows in agentic systems. When faced with complex, multi-step tasks, most agents either follow a pre-hardcoded pipeline or proceed without a clear, auditable execution plan. This plugin fundamentally changes that by teaching the agent to dynamically synthesize a task-specific workflow at runtime. Its core capability is “plan-first, dynamic orchestration”: the agent analyzes the request, selects an appropriate structural pattern (like sequential chaining, parallel processing, or routing), and then writes a structured, verifiable plan which is durably recorded before any work begins. The actual execution then relies solely on the harness’s battle-tested primitives (like sub-agents and tools), ensuring nothing extra breaks. This approach delivers explicit, versioned, and resumable workflows without adding runtime complexity. It is ideal for users running important multi-step work—such as code audits, coordinated research, or complex reports—who need their process to be transparent, reliable, and traceable, rather than left to the agent’s implicit behavior.
Screenshots
Use Cases
- When handling complex tasks that require clear phased execution
- When coordinating multiple sub-agents for parallel or divided collaboration
- When task requirements change dynamically, requiring real-time plan adjustment
Best For
- Developers or researchers running critical multi-step workflows with DeepSeek Harness
- Teams that need explicit, versioned, and traceable workflow plans
- Automation users who prioritize transparency and reliability over hardcoded pipelines
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.