Skip to main content
Glama

MCP Orchestration Server

flow.md2.12 kB
This is the working of the entire project: 1)User Input (via CLI or API) 2)MCP Adapter receives the query 3)Based on method and target: 4)Calls calendar_agent for reminders 5)Calls real_gmail_agent to send mail 6)Calls math_agent for math problems 7)Agent processes message using its handle_process() 8)Results returned to user or pushed to other services *base_agent.py* Base class for all agents. Handles: Logging Capability declaration Message processing Agent-to-agent communication Defines: MCPMessage: Communication format AgentCapability: Metadata for what each agent can do *calendar_agent.py* Parses and manages calendar events and reminders Handles queries like: “Remind me to call at 3 PM” “Schedule meeting tomorrow” Stores events in in-memory lists *real_gmail_agent.py* Sends actual emails using Gmail SMTP Uses: .env credentials (GMAIL_EMAIL, GMAIL_APP_PASSWORD) Professional email templates Can be used in workflow pipelines to notify users *gmail_agent.py* More advanced Gmail handler using Gmail API (via google-api-python-client) Can: Read inbox Send templated emails Search messages *math_agent.py* Solves math expressions and word problems Understands natural language like: “What is 15% of 300?” “Calculate area of a circle with radius 5” Uses eval safely for math expressions *mcp_adapter.py* Routes incoming messages from pipelines to correct agents Can call methods like process, send_email, info, etc. *mongodb_utils.py, mongodb_connection.py* Handle MongoDB connection Provide: Safe inserts Reconnection logic Index creation *file_utils.py, ocr_utils.py* Handle file reading, cleaning Perform OCR on documents, especially PDFs and images *uploads/* Temporary storage for uploaded files such as: PDFs Images Signature photos workflows/ Pipelines that connect multiple agents to accomplish tasks. *archive_search_agent.py* Reads PDF or scanned docs Extracts text via OCR Runs document summarization or question answering *live_data_agent.py* Pulls real-time data (e.g., weather, news) Useful for generating summaries or alerts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Nisarg-123-web/MCP2'

If you have feedback or need assistance with the MCP directory API, please join our Discord server