One-click install
npx skills add https://github.com/gitarbor/gitarbor-tui --skill opencode-agents-gitarbor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: opencode-agents
Source: https://github.com/gitarbor/gitarbor-tui/tree/main/.opencode/skills/opencode-agents
Command: npx skills add https://github.com/gitarbor/gitarbor-tui --skill opencode-agents-gitarbor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

OpenCode users face friction in creating and configuring agents across projects. This skill provides a structured, reusable blueprint to create and configure OpenCode agents (primary and subagents), assign prompts, tools, permissions, and models, and manage templates across global and project scopes.

Core Features & Use Cases

  • Agent creation and configuration: Create Markdown or JSON agent definitions with frontmatter and body instructions.
  • Templates and references: Reuse provided assets/templates and references/REFERENCE.md to standardize agent behavior.
  • Global vs project scope: Manage global and project-specific agent configurations via SKILL.md frontmatter and resources.
  • Permission and tool control: Configure tools, permission policies, and task orchestration to govern agent actions.
  • Templates for common roles: Use built-in templates like read-only reviewer, security auditor, docs writer, careful dev, etc.
  • Guidance and best practices: Step-by-step instructions for building and testing agents.

Quick Start

  • Review the skill's SKILL.md to understand the allowed agent definitions.
  • Place new agent definitions in .opencode/agents/NAME.md (or ~/.config/opencode/agents/NAME.md).
  • Reuse templates from assets/templates to bootstrap new agents.

Frequently Asked Questions about opencode-agents

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

FAQPage Schema
How do I configure OpenCode agents for specific project tasks?

Configure OpenCode agents by creating Markdown or JSON definition files with frontmatter, customizing prompts, tools, and permissions to tailor AI workflows for specific project tasks.

What is the difference between global and project-specific OpenCode agent configurations?

Global agent configurations apply across all projects via ~/.config/opencode/agents/, while project-specific configurations are placed in .opencode/agents/ to isolate agent behavior per project scope.

Can I reuse templates to bootstrap new OpenCode agents quickly?

Yes, you can reuse provided templates from the assets directory to bootstrap new agents, standardizing behavior for common roles like read-only reviewer, security auditor, and docs writer.

Do I need frontmatter to define an OpenCode agent in SKILL.md?

Yes, frontmatter is required in SKILL.md to define the agent's name and description, while optional references and assets directories provide supplementary documentation and templates.

How do I manage permissions and tool access for OpenCode subagents?

Manage permissions and tool access for OpenCode subagents by configuring permission policies and tool settings within the agent definition files to govern allowed actions and task orchestration.

What's the best way to structure agent definitions for a careful development workflow?

Structure agent definitions using Markdown files with frontmatter and body instructions, leveraging built-in templates like the careful dev template to enforce step-by-step development workflows.