agent-creator-opencode

Create and configure OpenCode CLI agents via Markdown files and JSON configurations.

Updated Nov 15, 2025
One-click install
npx skills add https://github.com/CowboyLogic/ai-dev --skill agent-creator-opencode
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-creator-opencode
Source: https://github.com/CowboyLogic/ai-dev/tree/main/skills/agent-creator-opencode
Command: npx skills add https://github.com/CowboyLogic/ai-dev --skill agent-creator-opencode

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill guides teams to create and configure OpenCode agents for the CLI, covering Markdown-based agent files, JSON configurations, and a robust permissions model to streamline agent deployment and governance.

Core Features & Use Cases

  • Markdown agent files in .opencode/agents for project-scoped agents
  • JSON config in opencode.json for global and per-agent customization
  • Permission model and orchestration guidance to enable safe delegation and multi-agent workflows
  • Use case: quickly scaffold an orchestrator that coordinates subagents during a large feature rollout

Quick Start

Run opencode agent create to scaffold a new agent, then edit .opencode/agents/<name>.md to customize the prompt, permissions, and model settings.

Frequently Asked Questions about agent-creator-opencode

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

FAQPage Schema
How do I create and configure OpenCode agents for the CLI?

To create OpenCode agents, run the agent creation command to scaffold a Markdown file, then edit the file in .opencode/agents to define the description, mode, model, prompts, and permissions for your CLI workflows.

What is the difference between project-scoped and global agents in OpenCode?

Project-scoped agents are stored as Markdown files in the .opencode/agents directory for localized workflows, whereas global agents are configured using the opencode.json file to enable consistent agent behavior across multiple projects.

How do I set up a permission model for multi-agent workflows in OpenCode?

You set up the permission model by defining permission rules within the Markdown agent files or JSON configurations, ensuring safe delegation and orchestration when coordinating subagents during complex feature rollouts.

Can I use JSON configurations to customize OpenCode agent behavior?

Yes, you can use JSON configurations in the opencode.json file to apply global and per-agent customization, allowing you to specify model settings, deployment scope, and orchestration guidance for your agents.

What's the best way to scaffold an orchestrator that coordinates subagents?

The best way to scaffold an orchestrator is using the agent creation command for step-by-step prompts, then customizing the generated Markdown file to define how the primary agent delegates tasks to subagents safely.

Do I need Markdown files or JSON configurations to define agent permissions in OpenCode?

You can define agent permissions using either Markdown files in .opencode/agents or JSON configurations in opencode.json, with both formats supporting the permission model needed for safe agent deployment and governance.