| name | yolo-auto |
| description | YOLO (You Only Live Once) full-stack automation skill with OpenClaw MCP bridge and GPU model selection. |
YOLO Auto Agent Skill
Overview
YOLO (You Only Live Once) full-stack automation skill with OpenClaw MCP bridge and GPU model selection. Enables autonomous multi-agent orchestration with intelligent task distribution across different GPU models.
Capabilities
- GPU Model Selection: Select optimal GPU model (claude-3-opus, gpt-4, gpu-5.3-codex)
- Multi-Agent Orchestration: Coordinate multiple agents for complex tasks
- OpenClaw Integration: Use OpenClaw MCP bridge for tool access
- Task Distribution: Distribute work across agents intelligently
- Result Aggregation: Collect and synthesize results from multiple sources
- Autonomous Execution: Execute complex workflows with minimal input
Tools Required
MCP Tools
yolo_select_gpu - Select GPU model for task
yolo_distribute_task - Distribute task to agents
yolo_aggregate_results - Aggregate results from agents
yolo_execute_workflow - Execute multi-step workflow
yolo_monitor_progress - Monitor workflow progress
openclaw_execute - Execute via OpenClaw bridge
File System Access
- Read: Full project access
- Write:
./yolo-workflows/, ./artifacts/, ./results/
Network Access
- None required - Local OpenClaw operations
Shell Commands
openclaw - OpenClaw CLI
python - Python for orchestration scripts
git - Version control
Usage Examples
Basic Workflow
codex $yolo-auto "Execute full-stack development workflow"
GPU Selection
codex $yolo-auto "Select claude-3-opus for complex reasoning task"
Multi-Agent Task
codex $yolo-auto "Distribute tasks: 3 agents for frontend, 2 for backend"
YOLO Architecture
┌─────────────────────────────────────────┐
│ YOLO Orchestrator │
├─────────────────────────────────────────┤
│ ┌─────────────┐ ┌─────────────┐ │
│ │ GPU Router │───▶│ Task Queue │ │
│ └─────────────┘ └─────────────┘ │
│ │ │ │
│ ▼ ▼ │
│ ┌─────────────┐ ┌─────────────┐ │
│ │ Agent Pool │◀───│ Work Stealer │ │
│ │ - claude-3 │ └─────────────┘ │
│ │ - gpt-4 │ │
│ │ - gpu-5.3 │ │
│ └─────────────┘ │
│ │ │
│ ▼ │
│ ┌─────────────┐ │
│ │ Result │ │
│ │ Aggregator │ │
│ └─────────────┘ │
└─────────────────────────────────────────┘
GPU Models
| Model | Strengths | Use Cases |
|---|
| claude-3-opus | Complex reasoning, coding | Architecture, planning |
| gpt-4 | Fast iteration, creativity | Prototyping, exploration |
| gpu-5.3-codex | Local processing | Security, privacy tasks |
OpenClaw Integration
OpenClaw provides MCP bridge for external tools:
{
"mcpServers": {
"openclaw": {
"command": "openclaw",
"args": ["serve"]
}
}
}
Workflow Definition
name: full-stack-web
stages:
- name: frontend
agents: 2
gpu: claude-3-opus
tasks:
- Create React components
- Implement state management
- name: backend
agents: 2
gpu: gpt-4
tasks:
- Design API schema
- Implement endpoints
- name: integration
agents: 1
gpu: claude-3-opus
tasks:
- Connect frontend to backend
- Test end-to-end
Output Format
The yolo-auto agent provides:
- Workflow execution reports
- Agent task distributions
- GPU utilization metrics
- Result aggregations
- Performance statistics
References
Installation: $ codex $skill-install https://github.com/zapabob/codex-yolo-auto-skill
Version: 1.0.0
Compatibility: Codex v2.14.0+