Available Servers
MCP Gateway supports many different MCP servers. Here are some of the most popular ones:
Search & Web
Section titled “Search & Web”- DuckDuckGo: Web search without API keys
- Brave: Web search using Brave Search API
- Exa: Neural search API
- ArXiv: Search and retrieve ArXiv papers
- Wikipedia: Search Wikipedia articles
Databases
Section titled “Databases”- PostgreSQL: Connect to PostgreSQL databases
- MySQL: Connect to MySQL databases
- MongoDB: Connect to MongoDB databases
- Redis: Connect to Redis instances
- Chroma: Vector database operations
- Pinecone: Vector database operations
Cloud Services
Section titled “Cloud Services”- AWS: AWS service integration
- Azure: Azure service integration
- GCP: Google Cloud Platform integration
- Vercel: Vercel deployment and management
- Heroku: Heroku app management
Development Tools
Section titled “Development Tools”- GitHub: GitHub repository operations
- GitLab: GitLab repository operations
- Code Interpreter: Execute code in sandbox
- Filesystem: File system operations
AI Services
Section titled “AI Services”- OpenAI: OpenAI API integration
- Anthropic: Anthropic Claude API
- Hugging Face: Hugging Face models
- ElevenLabs: Text-to-speech service
Communication
Section titled “Communication”- Slack: Slack workspace integration
- Discord: Discord bot operations
- Gmail: Gmail API integration
- Notion: Notion workspace integration
Configuration Examples
Section titled “Configuration Examples”Here are some common server configurations:
DuckDuckGo (No configuration needed)
Section titled “DuckDuckGo (No configuration needed)”from agentbox import Sandbox
sandbox = Sandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"duckduckgo": {}
}
)from agentbox import AsyncSandbox
sandbox = await AsyncSandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"duckduckgo": {}
}
)ArXiv (With storage path)
Section titled “ArXiv (With storage path)”from agentbox import Sandbox
sandbox = Sandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"arxiv": {
"storagePath": "/tmp/arxiv"
}
}
)from agentbox import AsyncSandbox
sandbox = await AsyncSandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"arxiv": {
"storagePath": "/tmp/arxiv"
}
}
)GitHub (With authentication)
Section titled “GitHub (With authentication)”from agentbox import Sandbox
sandbox = Sandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"github": {
"token": "your-github-token",
"owner": "your-username",
"repo": "your-repo"
}
}
)from agentbox import AsyncSandbox
sandbox = await AsyncSandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"github": {
"token": "your-github-token",
"owner": "your-username",
"repo": "your-repo"
}
}
)PostgreSQL (With connection string)
Section titled “PostgreSQL (With connection string)”from agentbox import Sandbox
sandbox = Sandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"postgres": {
"connectionString": "postgresql://user:pass@host:5432/db"
}
}
)from agentbox import AsyncSandbox
sandbox = await AsyncSandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"postgres": {
"connectionString": "postgresql://user:pass@host:5432/db"
}
}
)Multiple Servers
Section titled “Multiple Servers”You can configure multiple servers in a single sandbox:
from agentbox import Sandbox
sandbox = Sandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"duckduckgo": {},
"arxiv": {"storagePath": "/tmp/arxiv"},
"context7": {},
"github": {
"token": "your-github-token",
"owner": "your-username",
"repo": "your-repo"
}
}
)from agentbox import AsyncSandbox
sandbox = await AsyncSandbox.create(
api_key="ab_xxxxxxxxxxxxxxxxxxxxxxxx",
mcp={
"duckduckgo": {},
"arxiv": {"storagePath": "/tmp/arxiv"},
"context7": {},
"github": {
"token": "your-github-token",
"owner": "your-username",
"repo": "your-repo"
}
}
)