Agent Team Setup Skill

A reusable opencode skill that scaffolds a complete team of AI agents in any project.

Prerequisites

Before using this skill, you need to have OpenCode CLI installed.

Installing OpenCode CLI

Quick Install (Recommended)

curl -fsSL https://opencode.ai/install | bash

Alternative Installation Methods

Using npm:

npm install -g opencode-ai

Using Homebrew (macOS/Linux):

brew install anomalyco/tap/opencode

Using Chocolatey (Windows):

choco install opencode

Using Scoop (Windows):

scoop install opencode

Using Arch Linux:

sudo pacman -S opencode           # Stable
paru -S opencode-bin              # Latest from AUR

Using Mise:

mise use -g github:anomalyco/opencode

Using Docker:

docker run -it --rm ghcr.io/anomalyco/opencode

Verify Installation

opencode --version

You should see the version number (e.g., 1.14.33).

Choose a Provider

OpenCode + DeepSeek V4 Flash

New agent teams use OpenCode with the free DeepSeek V4 Flash model by default:

opencode/deepseek-v4-flash-free

Start OpenCode and run /connect, then select OpenCode (OpenCode Zen). Use /models to confirm or change the selected model. Credentials are managed by OpenCode and are not written to generated project files.

Other Providers

Provider	Setup	Model ID example
Cloudflare Workers AI	/connect → Cloudflare Workers AI	cloudflare-workers-ai/@cf/zai-org/glm-5.2
ChatGPT/OpenAI	/connect → OpenAI, then /models	openai/<model-id>
Another connected provider	Select it with /connect, then choose a model with /models	Provider-specific model ID

ChatGPT login options, including free or paid plans, depend on the OpenCode version and the account. OpenCode will show the available options during /connect.

---

What It Does

This skill helps you create production-ready agent teams for automating multi-step workflows. When you load this skill and describe your pipeline, it will:

1. Plan the pipeline (agents, handoffs, tools)
2. Scaffold the .opencode/ directory structure
3. Generate the opencode.json coordinator config
4. Create subagent .md files with proper permissions
5. Create skill .md files for domain knowledge
6. Create custom tool .py + .ts pairs
7. Create the AGENTS.md project instructions file (or use /init)
8. Create custom commands for repetitive workflows (optional)
9. Test the pipeline end-to-end

---

Installing This Skill

Option 1: Clone from GitHub (Recommended)

# For global installation (available in all projects)
git clone https://github.com/bighadj22/opencode-skill.git ~/.config/opencode/skills/agent-team-setup

# Or for .agents directory
git clone https://github.com/bighadj22/opencode-skill.git ~/.agents/skills/agent-team-setup

Option 2: Manual Installation

1. Download or clone this repository
2. Copy the folder to your OpenCode skills directory:

# For global installation
cp -r agent-team-setup-skill ~/.config/opencode/skills/agent-team-setup

# Or for .agents directory
cp -r agent-team-setup-skill ~/.agents/skills/agent-team-setup

Option 3: Project-Specific Installation

# Install for a specific project only
cd /path/to/your/project
mkdir -p .opencode/skills
git clone https://github.com/bighadj22/opencode-skill.git .opencode/skills/agent-team-setup

Verify Installation

After installation, restart OpenCode and verify the skill is available:

opencode
# Then in OpenCode, type:
# "What skills do you have?" or "List available skills"

You should see agent-team-setup in the list of available skills.

How to Use

Usage

In any project, load the skill and describe your workflow:

@skill agent-team-setup

Set up an agent team for my SaaS marketing blog:
research trending topics → write SEO articles → generate social media images → publish to WordPress

The skill will guide you through the entire setup process and help you make decisions about:

• Agent roles and responsibilities
• Handoff mechanisms between agents
• Custom tools needed
• Model selection and temperature settings
• Permission configurations

What's Inside

agent-team-setup-skill/
  SKILL.md              # The full skill instructions (loaded by opencode)
  README.md             # This file
  references/           # Comprehensive OpenCode documentation references
    opencode-agents.md
    opencode-skills.md
    opencode-tools.md
    opencode-commands.md
    opencode-rules.md
    opencode-config.md
    opencode-permissions.md
    opencode-mcp.md
    opencode-builtin-tools.md
  templates/
    opencode.json       # Coordinator config template
    AGENTS.md           # Project instructions template
    agent.md            # Subagent definition template
    skill.md            # Skill definition template
    tool.py             # Python script template
    tool.ts             # TypeScript wrapper template
    package.json        # .opencode/ package.json
    gitignore           # .opencode/.gitignore
    requirements.txt    # Python deps template

Use Cases

This skill works for any multi-step automated workflow:

• Content Creation: Research → Write → Edit → Publish → Promote
• Software Development: Analyze → Design → Code → Test → Deploy
• Data Analysis: Collect → Clean → Analyze → Visualize → Report
• Customer Support: Triage → Investigate → Respond → Follow-up
• Marketing: Research → Plan → Create → Schedule → Track

Features

• Provider Flexibility: Use OpenCode, Cloudflare Workers AI, ChatGPT/OpenAI, or another provider through /connect
• Complete Setup: Creates all necessary files and directories
• Best Practices: Follows OpenCode conventions and patterns
• Comprehensive Documentation: Includes 9 detailed reference files covering all OpenCode features
• Flexible Architecture: Adapts to any domain or workflow
• Permission Control: Proper security boundaries for each agent
• Custom Tools: Templates for Python + TypeScript tool integration
• Skills System: Knowledge documents for domain-specific guidance
• Commands System: Reusable prompt templates for common workflows
• AGENTS.md Integration: Project-level instructions with /init support

License

Apache-2.0