MCP_Servers_SoulfulAI

SoulfulAI's MCP_Servers project provides a suite of context-aware MCP servers that integrate with real-world services. Each server exposes two-way context access, enabling an LLM-driven client to query or push data to the connected service (e.g., Firebase, Salesforce, OBS, or Supabase) while maintaining a unified MCP workflow. You can run multiple servers in parallel and route prompts to the appropriate service handler. The included tests and integration points help verify end-to-end flows across dynamic context switches, real-time data ops, and CRM interactions. To experiment, deploy the servers locally, connect an MCP client, and explore prompts that trigger reads, writes, or state changes across the integrated services.

Prerequisites:

• Node.js (v14+ or newer) and npm
• Git

Step-by-step installation:

1. Clone the repository git clone https://github.com/iamRajamurugan/MCP_Servers_SoulfulAI.git cd MCP_Servers_SoulfulAI

2. Install dependencies (per-server setup is in the adya_mcp_hackathon/js/server directory and may require individual dependency installs): npm install

3. Configure environment variables:

   • Create a .env file or export vars as needed for each server, corresponding to the env keys listed in the mcp_config block above. Example: export FIREBASE_CONFIG='{"projectId":"..."}' export SF_CLIENT_ID='...' export SF_CLIENT_SECRET='...' export SF_INSTANCE_URL='https://your-instance.salesforce.com'

4. Start the servers (example for Context7 server only): node adya_mcp_hackathon/js/server/context7.js &

   You can similarly start the other servers by running their respective entry files. Monitor logs to ensure successful connections to the target services.

5. Run tests (if desired): npm test

Tips:

• Ensure that API keys and service credentials are kept securely (use environment variables or a secret manager).
• Some services may require additional setup (e.g., Firebase emulators, OBS WebSocket permissions, or Salesforce OAuth flows).
• The MCP servers are designed to be run together; consider orchestrating with a process manager (e.g., PM2) or a docker-compose setup for consistent startup order.
• If you modify any server code, re-run tests to confirm that critical paths remain covered.
• For troubleshooting, verify network access from the host to each external service and check the MCP client logs for prompt routing decisions.