We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/0ui-labs/agentic-workbench'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
quickstart.md•1.24 kB
# Quick Start
This guide will help you run your first task with Agentic Workbench.
## 1. Set Up Your Catalog
Create a catalog directory with your tools:
```bash
mkdir -p catalog/communication/slack
```
Create a category file `catalog/communication/_category.yaml`:
```yaml
name: Communication
description: Tools for messaging and communication
```
Create a service file `catalog/communication/slack/_service.yaml`:
```yaml
name: Slack
description: Slack messaging platform
mcp_server: slack-mcp
```
## 2. Configure Environment
Create a `.env` file:
```bash
cp .env.example .env
```
Edit `.env` with your LLM provider settings:
```env
AWB_LLM_PROVIDER=cerebras
AWB_LLM_API_KEY=your-api-key
```
## 3. Run a Task via CLI
```bash
awb run "Send a message to #general saying hello"
```
## 4. Start as MCP Server
```bash
awb serve --catalog ./catalog
```
## 5. Integrate with Claude Code
Add to your `.mcp.json`:
```json
{
"mcpServers": {
"agentic-workbench": {
"command": "awb",
"args": ["serve", "--catalog", "./catalog"]
}
}
}
```
## Next Steps
- [Configuration](configuration.md) - Learn about all configuration options
- [Catalog Structure](../concepts/catalog.md) - Understand the catalog format