mcp s

This MCP server collection aggregates a suite of MCP (Model Context Protocol) servers maintained by Dunia Labs. Each subdirectory within the repository corresponds to a specific integration (e.g., Google Drive, PostgreSQL, Notion, Google Calendar, Figma, GitHub, Zendesk, Skills, Canva, Stripe, and more) and provides a ready-to-run MCP service with STDIO transport and Docker support. When integrated with Peta Core, these MCPs offer secure credential handling, multi-tenant support, granular permissions, and human-in-the-loop approval workflows for risky operations. You can connect any of the individual MCP servers to your Peta Core deployment or run them independently as needed.

To use, deploy the desired MCP servers (preferably via Docker for portability) and connect them to your Peta Core instance using the MCP protocol. Each server advertises a suite of tools (e.g., file operations for Drive, database queries for PostgreSQL, repository and issue management for GitHub, calendar and event controls for Google Calendar, etc.). Through Peta Core, you can invoke these tools through the standard MCP interface, authorize access via tokens or credentials stored in a vault, and enforce per-tenant or per-key permissions. The configuration supports runtime token refresh and STDIO transport for efficient local or cloud-based deployment.

Prerequisites:

• Docker installed on your host or deployment environment
• Access to a secret vault (optional but recommended) for credentials
• Basic knowledge of MCP and Peta Core integration

Step-by-step:

1. Pull the MCP servers collection (Docker image): docker pull dunialabs/mcp-servers:latest

2. Run the collection container (adjust image tag as needed): docker run -d --name dunialabs-mcp-servers -i dunialabs/mcp-servers:latest

3. Configure vault and connection details:

   • If you are using a vault, ensure VAULT_URL and credentials are available to the container
   • Set PETA_CORE_API_ENDPOINT to point to your Peta Core instance

4. Integrate with Peta Core:

   • In your Peta Core setup, register the MCP servers from the collection using the provided MCP protocol configuration
   • Ensure proper token-based authentication and per-tenant permissions are in place

5. Verify operation:

   • Access the admin or status endpoints (as documented in each sub-server) or test with a sample MCP tool invocation from Peta Core

Optional:

• Build from source if you maintain the collection yourself, following the individual sub-server READMEs for toolsets and configuration options.

Tips and common considerations:

• Each sub-server supports Docker as a deployment option; prefer using Docker to simplify environment parity across development and production.
• Many servers offer token-based authentication and runtime token refresh; ensure your Peta Core tokens are refreshed before expiry to avoid interruptions.
• If you’re integrating with Google services (Drive, Calendar), ensure you configure OAuth scopes and any required refresh tokens in your vault or environment.
• For REST-based or configurable servers (e.g., MCP REST Gateway), you can auto-generate config from OpenAPI specs to accelerate setup.
• When deploying multiple MCP servers, consider grouping them by function (data access, collaboration, payments, etc.) and manage per-tenant access controls to limit tool usage per API key.
• If encountering issues, check STDIO transport logs for real-time tool invocations and enable verbose logging in the environment variables if supported by the sub-servers.