skills.rest
dingdyan
dingdyanCOMMUNITY
Explore all skills by this author →
GitHub

Agent Company (代理公司) - Hierarchical Agent Management

Orchestrate a Manager–Worker hierarchy to decompose tasks and synthesize deliverables.

Updated Apr 5, 2026
Search Tags:#code review#research#code generation#project planning#parallel execution#agent orchestration#manager-worker
One-click install
npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill agent-company-hierarchical-agent-management
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Agent Company (代理公司) - Hierarchical Agent Management
Source: https://github.com/dingdyan/openclaw-workspace-v2/tree/main/skills/agent-company
Command: npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill agent-company-hierarchical-agent-management

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Complex requests often require multiple skills at once (research, implementation, and verification), but doing that coordination manually is slow and error-prone.

Core Features & Use Cases

  • Manager–Worker orchestration: Decomposes a high-level request into subtasks, assigns work to specialized agents, and consolidates outputs.
  • Role-based execution pipeline: Uses dedicated roles for Research, Coding, and Review to improve quality and reduce rework.
  • Project plan driven workflow: Creates and updates a per-project PLAN.md to track progress and decide what to execute next (including optional review steps).

Quick Start

Ask the AI to create a project plan and run the Manager to spawn Researcher and Coder agents for your task, then verify the result with a Reviewer before delivering the final report.

Frequently Asked Questions about Agent Company (代理公司) - Hierarchical Agent Management

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate multiple AI agents for parallel research and code generation?▼

You can orchestrate parallel research and code generation by using a Manager-Worker hierarchy that decomposes complex requests into subtasks, assigns them to specialized agents, and synthesizes the outputs into a final deliverable.

What is a manager-worker hierarchy for complex project planning and execution?▼

A manager-worker hierarchy is an agent orchestration structure where a Manager decomposes high-level tasks, delegates subtasks to specialized Worker agents, and consolidates their results into a verified final project deliverable.

How do I automate code review and quality verification after parallel code generation?▼

Automate code review by running an optional Reviewer agent after implementation, which evaluates the consolidated worker outputs to produce a clear pass or fail style review report for correctness and safety.

Can I use agent orchestration to manage end-to-end builds that require research and implementation?▼

Yes, agent orchestration is specifically suited for end-to-end builds, using dedicated roles for Research, Coding, and Review to coordinate parallel execution and reduce manual rework across the project pipeline.

How do I track progress when delegating subtasks to multiple specialized agents?▼

You can track progress by creating and updating a per-project PLAN.md file, which the Manager agent uses to monitor completed subtasks and decide what to execute next throughout the workflow.

Do I need to manually share results between subagents during parallel execution?▼

No, subagents share results automatically via files or tool outputs, allowing the Manager to consolidate research and implementation outputs seamlessly before the final review and delivery.

Related Skills

Explore More Product & Management Skills→
agentscope-aiagentscope-ai
5.3k

project-participation

Coordinate multi-worker project tasks using plan.md and git attribution.

Official
Intermediate
FrekiManagarmFrekiManagarm

Verification & Quality Assurance

Compute truth scores and run correctness, security, performance, and documentation checks with JSON reports and Git rollback.

Community
Advanced
scooter-lacroixscooter-lacroix
8

review

Orchestrate multi-agent code review with parallel specialized agents and synthesized feedback.

Community
Advanced
FlavioCFOliveiraFlavioCFOliveira

roadmap-coordinator

Coordinate task workflows via the Groadmap CLI with state transitions and delegation.

Community
Advanced