home / skills / frankxai / arcanea / arcanea-coding-agent

arcanea-coding-agent skill

This skill turns opencode into an Arcanean creative workspace with 64 specialized agents and real-time multi-agent collaboration.

npx playbooks add skill frankxai/arcanea --skill arcanea-coding-agent

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

2.3 KB

---
name: arcanea-coding-agent
description: Arcanea creative agent integration for opencode
version: 1.0.0
author: Arcanea Team
emoji: 🌟
---

# Arcanea Coding Agent Skill

## Overview

This skill transforms opencode into an Arcanean creative workspace, giving you access to 64 specialized agents for creative projects directly within your editor.

## Installation

```bash
# In opencode, the skill is automatically available when in Arcanea directory
# Or manually enable:
opencode skill enable arcanea-coding-agent
```

## Features

- **64 Specialized Agents** - Access all elemental courts
- **Inline Invocation** - @agent mentions in any file
- **Workflow Support** - Multi-agent orchestration
- **Context Awareness** - Agents know your current file/project
- **Real-time Streaming** - See agent responses as they generate
- **Cache & Learning** - Faster responses over time

## Usage

### Basic Agent Invocation

Type in any file or chat:
```
@ignition help me brainstorm ideas for this function
```

### Context-Aware Help

Select code, then invoke:
```
@refinement improve this selected code
```

### Multi-Agent Workflows

```
@orchestration create character profile for my protagonist
```

### Commands

| Command | Description |
|---------|-------------|
| `arcanea:invoke` | Invoke agent with current context |
| `arcanea:agents` | List all available agents |
| `arcanea:workflows` | Show available workflows |
| `arcanea:mode arcane` | Switch to mystical language |
| `arcanea:mode pro` | Switch to professional terms |

## Configuration

```json
{
  "arcanea": {
    "mode": "arcane",
    "defaultAgent": "orchestration",
    "aiProvider": "hybrid",
    "claudeApiKey": "${CLAUDE_API_KEY}",
    "cacheEnabled": true,
    "streaming": true
  }
}
```

## Key Agents for Coding

- **@structure** - System architecture and design
- **@optimization** - Performance improvements
- **@refinement** - Code polishing and cleanup
- **@clarity** - Documentation and explanation
- **@debug-exorcist** - Bug hunting and fixes

## Integration

This skill works with:
- opencode editor commands
- File context awareness
- Selection-based operations
- Project-wide knowledge

## Learn More

- Full Documentation: `AGENT_ARCHITECTURE_v4.md`
- Claude Instructions: `.claude/CLAUDE.md`
- UI Demo: `PREMIUM_UI_DEMO.html`

Overview

This skill turns your opencode workspace into an Arcanea creative coding environment with 64 specialized agents available inline. It brings multi-agent orchestration, context-aware assistance, and real-time streaming responses directly into your editor. You get a blend of creative frameworks and pragmatic coding tools to speed iteration and improve code quality.

How this skill works

Agents are invoked inline using @mentions or editor commands and run against the current file, selection, or project context. The system routes requests to the configured AI provider, streams responses as they generate, and caches learned patterns for faster follow-ups. Workflows let multiple agents coordinate tasks—e.g., drafting, refining, and testing—so complex creative or technical jobs execute end-to-end.

When to use it

Brainstorming architecture, features, or story seeds directly inside your codebase
Polishing and refactoring selected code with context-aware suggestions
Running multi-step creative workflows like character design or feature kickoff
Debugging tricky issues with a bug-hunting specialist agent
Generating documentation, explanations, or onboarding notes for project files

Best practices

Invoke agents inline with clear prompts and include selected code or a short project description
Use the orchestration workflow for multi-step tasks to keep task handoffs structured
Start with pro mode for technical precision, switch to arcane mode for creative language when exploring ideas
Enable caching and streaming for faster interactive sessions and smoother iteration
Review and adapt agent outputs; agents accelerate work but developer oversight ensures correctness

Example use cases

Ask @structure to produce a high-level system diagram and API surface for a new feature
Select a function and run @refinement to clean up style, add comments, and simplify logic
Use @debug-exorcist on failing tests to get stepwise hypotheses and suggested fixes
Run @orchestration to generate a protagonist profile, then feed it to content generators to draft scenes
List available agents with the command palette and quickly assign roles in a multi-agent workflow

FAQ

How do I invoke an agent inside a file?

Type an @mention with the agent name and a prompt, or use the arcanea:invoke command to call the default agent with current context.

Can agents access my whole project?

Yes—agents operate on the current file, selection, or project context depending on the invocation and permissions you grant.

What modes are available and when to switch?

Use 'pro' for concise, technical outputs and 'arcane' for creative, mythic language. Switch modes when you need either precision or stylistic exploration.