Agents Company for HappyCapy

AI Multi-Agent Orchestration System based on the ancient Chinese "Three Departments and Seven Ministries" governance model.

10 specialized AI Agents collaborate through a structured hierarchy: Strategy Hub plans, Review Hub audits, Execution Hub dispatches, and 7 Ministries handle domain-specific work -- all coordinated through a real-time dashboard.

OpenClaw Python Agents Dashboard License React

Quick Start

One-line install:

curl -fsSL https://raw.githubusercontent.com/rogerhorsley/agents_company_for_happycapy/main/setup.sh | bash

Or manually:

git clone https://github.com/rogerhorsley/agents_company_for_happycapy.git
cd agents_company_for_happycapy
chmod +x install.sh && ./install.sh

Start the dashboard:

python3 dashboard/server.py
# Open http://127.0.0.1:7891

Architecture

                    +------------------+
                    |   Strategy Hub   |  Plan & Research
                    |     (策枢)       |
                    +--------+---------+
                             |
                    +--------v---------+
                    |   Review Hub     |  Audit & Approve
                    |     (衡枢)       |
                    +--------+---------+
                             |
                    +--------v---------+
                    |  Execution Hub   |  Dispatch & Monitor
                    |     (行枢)       |
                    +---+----+----+----+
                        |    |    |
          +-------------+    |    +-------------+
          |                  |                  |
   +------v------+   +------v------+   +-------v-----+
   | Research (谋)|   |  HR (人)    |   | Finance (财) |
   | Brand (品)   |   | Security(安)|   | Compliance(规)|
   | Tech (技)    |   |             |   |              |
   +-------------+   +-------------+   +--------------+

10 Agents

Features

Dashboard (Real-time Control Panel)

• Task Board -- Task cards with state tracking (Draft/Review/Doing/Done)
• Department Monitor -- Live agent status and workload
• Court Discussion -- Multi-agent collaborative deliberation
• Official Overview -- Agent profiles and performance stats
• Model Config -- Switch LLM models per agent
• Skills Config -- Manage agent skills and tools
• Sessions -- View and manage sub-tasks
• Memorial Hall -- Archived completed tasks
• Template Library -- Reusable task templates
• Morning Brief -- Daily intelligence summary via Feishu

System Setup (New)

• OpenClaw Detection -- Auto-detect local OpenClaw installation, inherit memory/config
• Channel Config -- Feishu webhook management (add/edit/test/delete channels)
• HappyCapy Login -- User authentication with personal API keys and model selection

Screenshots

Task Board	Dispatch	Court Discussion

Hub Monitor	Official Overview	Model Config

Skills Config	Sessions	Memorials

Templates	Morning Brief	System Setup

Tech Stack

Project Structure

agents_company_for_happycapy/
  agents/              # 10 agent SOUL.md definitions
  dashboard/           # Backend server + built frontend
    server.py          # API server (stdlib HTTP)
    dist/              # Built React frontend
  app/
    frontend/          # React 18 source (TypeScript)
    backend/           # FastAPI backend (optional)
    migration/         # Database migration scripts
    scripts/           # Internal utilities
  scripts/             # Sync, refresh, and automation scripts
  data/                # Runtime data (gitignored)
  docs/                # Documentation and screenshots
  install.sh           # Local installer
  setup.sh             # Remote one-line installer

Configuration

Environment Variables

CLI Shortcuts (after install)

ac-dashboard     # Start dashboard on port 7891
ac-sync          # Start background data sync loop
ac-update        # Pull latest and reinstall

Docker

docker compose up
# or
docker run -p 7891:7891 rogerhorsley/agents_company_for_happycapy

Contributing

See CONTRIBUTING.md for guidelines.

License

MIT License. See LICENSE.

_{Based on the open source project cft0808/edict.}