Welcome to the Agentic Edge Functions repository by the Agentics Foundation. This open-source project serves as the technological foundation for distributed autonomous AI agents operating at the edge of networks.
Agentic Edge Functions combine serverless edge computing with autonomous AI capabilities. Running at the network edge, these functions enable AI agents to:
- Operate independently near users for lower latency
- Communicate securely with other agents and systems
- Access controlled resources via defined interfaces
- Scale dynamically based on workload requirements
This infrastructure powers the Agentic Dashboard and Agentic Inbox systems at agentics.org/dashboard and is being actively developed in the open, welcoming community contributions.
A powerful Model Context Protocol server with advanced AI capabilities by the Agentics Foundation. Built on the OpenAI Agents API/SDK using TypeScript, this package implements a comprehensive MCP server that enhances AI agents with sophisticated tools and orchestration capabilities:
- 🔍 Web Search Research: Generate comprehensive reports with up-to-date information from the web using
gpt-4o-search-preview - 📝 Smart Summarization: Create concise, well-structured summaries with key points and citations
- 🗄️ Database Integration: Query and analyze data from Supabase databases with structured results
- 👥 Customer Support: Handle inquiries and provide assistance with natural language understanding
- 🔄 Agent Orchestration: Seamlessly transfer control between specialized agents based on query needs
- 🔀 Multi-Agent Workflows: Create complex agent networks with parent-child relationships and shared context
- 🧠 Context Management: Sophisticated state tracking with memory, resources, and workflow management
- 🛡️ Guardrails System: Configurable input and output validation to ensure safe and appropriate responses
- 📊 Tracing & Debugging: Comprehensive logging and debugging capabilities for development
- 🔌 Edge Function Deployment: Ready for deployment as Supabase Edge Functions
- 🔄 Streaming Support: Real-time streaming responses for interactive applications
# Install globally
npm install -g @agentics.org/agentic-mcp
# Or as a project dependency
npm install @agentics.org/agentic-mcpThis repository contains edge functions implementing various patterns for autonomous agent deployment and operation. Built on Supabase's serverless infrastructure and Deno runtime, these functions enable scalable, efficient, and globally distributed agent capabilities.
Additionally, it includes the scripts/agentic-mcp/ package, a powerful Model Context Protocol server that enhances AI agents with advanced tools and orchestration capabilities.
Edge functions are serverless functions running at network edges, providing low-latency, high-availability compute without server management. Supabase Edge Functions use Deno, a modern secure JavaScript/TypeScript runtime with built-in security features.
Deno is a modern, secure runtime for JavaScript and TypeScript that provides several advantages:
- Security-first design: Explicit permissions for file, network, and environment access
- Built-in TypeScript support: No configuration needed
- Modern JavaScript: Support for ES modules, top-level await, and more
- Standard library: Comprehensive, reviewed standard modules
- Single executable: No dependency management headaches
Our edge functions fall into several categories:
- agent_alpha - Basic agent implementation with ReAct pattern
- agent_beta - Enhanced agent with additional capabilities
- agent_stream - Streaming response agent
- agent_websocket - WebSocket-based agent communication
- agentic_inbox_agent - Agent for managing message inboxes
- agent-manager - Orchestration and management of multiple agents
- edge_deployment - Tools for deploying and managing edge functions
- resend - Email communication using Resend API
- send-contact-notification - Contact form notifications
- github-api - GitHub API integration
- mcp-server - Model Context Protocol server implementation
- stripe_check-subscription-status - Subscription status verification
- stripe_create-portal-session - Stripe customer portal integration
- feedback - User feedback collection
- meta-function-generator - Dynamic function generation
- hello-world - Simple example function
- test-function - Testing utilities
- Real-time Communication: Agents can communicate in real-time using Supabase's real-time channels
- Database Integration: Seamless integration with Supabase PostgreSQL database
- Secure Secrets Management: Environment variables for secure credential management
- Scalable Architecture: Functions scale automatically with demand
- Global Distribution: Deploy functions globally for low-latency responses
To get started with these edge functions:
- Clone this repository
- Set up a Supabase project
- Configure environment variables
- Deploy functions to your Supabase project
For beginners, we recommend following our step-by-step tutorials to learn how to create, deploy, and test edge functions. For detailed instructions, see the Getting Started Guide.
Comprehensive documentation is available in the docs directory:
- Function Documentation - Detailed documentation for each function
- Supabase Edge Functions - Overview of Supabase Edge Functions
- Real-time Channels - Using Supabase real-time channels
- Environment Variables - Comprehensive list of all environment variables
- Secrets Management - Best practices for managing secrets
- Database Triggers - Using database triggers with edge functions
- Deployment Guide - Deploying edge functions to production
- Tutorials Overview - Beginner-friendly guides for Agentic Edge Functions
- Creating Your First Edge Function - Learn the basics of edge functions
- Building a Basic Agentic Function - Create your first agentic edge function
- Deploying and Testing Edge Functions - Learn how to deploy and test your functions
Contributions are welcome! Please see our Contributing Guidelines for more information.
This project is licensed under the terms of the license included in the repository.
Created by rUv, Agentics Foundation founder.