# AGI-MCP Configuration Directory
This directory contains AGI-MCP server configurations, skills, agents, and validation hooks.
## Directory Structure
```
.agi-mcp/
├── skills/ # Skills for task orchestration
│ ├── plan-with-team.md # Team planning skill
│ ├── plan.md # Quick planning skill
│ ├── build.md # Plan implementation skill
│ ├── github.md # GitHub operations skill
│ └── playwright.md # Browser automation skill
├── agents/ # Specialized agent configurations
│ └── team/ # Team member definitions
└── hooks/ # Lifecycle hooks
└── validators/ # Validation scripts for skills
```
## Skills
Skills in `.agi-mcp/skills/` are automatically loaded by the AGI-MCP server.
### Available Skills
- **plan-with-team.md** - Creates detailed implementation plans with team orchestration
- **plan.md** - Quick engineering implementation plans
- **build.md** - Implements plans from specification files
- **github.md** - Comprehensive GitHub operations (30+ actions)
- **playwright.md** - Browser automation and web testing
## Agents
Team member agents can be defined in `.agi-mcp/agents/team/` and will be available for the plan-with-team skill to orchestrate.
## Hooks
### Validators
- **validate_new_file.py** - Checks if a new file was created in specified directory
- **validate_file_contains.py** - Validates that files contain required sections
These validators are used by the plan-with-team skill to ensure plan documents meet specifications.
## Usage
The plan-with-team skill can be executed via MCP:
```typescript
await tools.handleToolCall('execute_skill', {
skill_name: 'plan-with-team',
context: {
user_prompt: 'Implement user authentication system',
orchestration_prompt: 'Use security-focused approach'
}
});
```
## Compatibility
This is the native AGI-MCP configuration structure. All skills, agents, and hooks are centralized in this directory for maximum efficiency and maintainability.
