OpenCode Dedalus Skill

A unified skill for OpenCode that provides access to latest proprietary AI models through Dedalus AI API (GPT-5, Claude 4.5, Gemini 2.5, Grok 4).

Installation

Prompt OpenCode to set up this skill globally:

Install the dedalus skill globally in my machine based on https://github.com/raychrisgdp/opencode-skill-dedalus

OpenCode will clone the repository and set up the skill for you.

Setup

1. Get Your Dedalus API Key

Visit dedaluslabs.ai to sign up and get your API key.

2. Configure API Key

Once the skill is installed, create a .env file in the skill directory:

echo "DEDALUS_API_KEY=your_api_key_here" > ~/.config/opencode/skill/dedalus/.env
chmod 600 ~/.config/opencode/skill/dedalus/.env

3. Source Environment

Before using the skill, source the environment file:

source ~/.config/opencode/skill/dedalus/.env

Usage

Once installed, OpenCode agents can discover and load this skill automatically via the skill tool. The skill provides:

Available Models

Provider	Model	Best For
OpenAI	openai/gpt-5.2	General purpose, balanced performance
OpenAI	openai/o3	Complex reasoning
Anthropic	anthropic/claude-opus-4-5	Nuanced analysis, coding
Google	google/gemini-2.5-pro	Creative tasks
xAI	xai/grok-4	Up-to-date knowledge

Example API Call

curl --request POST \
  --url https://api.dedaluslabs.ai/v1/chat/completions \
  --header "Authorization: Bearer $DEDALUS_API_KEY" \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "openai/gpt-5.2",
    "messages": [
      {"role": "user", "content": "Hello! How are you?"}
    ]
  }'

Integration with OpenCode

This skill enables OpenCode to use Dedalus AI API for chat completions with OpenAI-compatible endpoints. The skill automatically loads your API key from the .env file when making requests.

Files

• SKILL.md - Full documentation and examples
• .env - Contains your API key (NOT committed to repo)
• .gitignore - Excludes sensitive files

Troubleshooting

• 401 Unauthorized: Check your API key is correct in .env
• 404 Not Found: Use format provider/model (e.g., openai/gpt-5.2)
• 429 Rate Limit: Reduce request frequency

Links

• Dedalus AI Documentation
• API Reference