15 AI terms related or similar to MCP (Model Context Protocol) and Agent
Written by - Millan Kaul
Let’s get into these most common yet important 15 today AI terms related or similar to MCP (Model Context Protocol) and Agentic AI.
| Term | Definition | Example Use | Reference Link |
|---|---|---|---|
| MCP (Model Context Protocol) | Protocol enabling AI agents to share structured context and communicate tool usage. | MCP lets AI agents coordinate by sharing what tools and memory they use for complex tasks. | Auth0 Blog |
| Agent | Autonomous AI component that acts independently to complete tasks using reasoning and learning. | An agent autonomously schedules meetings and sends invites without human input. | Perplexity Blog |
| A2A (Agent-to-Agent) | Communication protocol for AI agents to interact securely and collaboratively. | A2A allows a chatbot agent to pass a user query to a specialized knowledge agent for an accurate answer. | Google Blog |
| ACP (Agent Communication Protocol) | IBM-driven standard for structured messaging between AI agents and external systems. | ACP facilitates ticket updates and task assignments between customer service AI and human agents. | IBM Think Blog |
| AG-UI (Agent-User Interaction) | Protocol standardizing AI agent communication with user-facing applications for real-time interaction. | AG-UI supports chatbots sending live updates and receiving user inputs during a conversation. | IBM Think Blog |
| Agora | Communication protocol allowing LLM-based agents to negotiate protocols autonomously. | LLM agents in Agora negotiate to decide the best method for processing a request. | IBM Think Blog |
| ANP (Agent Network Protocol) | Network protocol framework for large-scale agent interactions over distributed systems. | ANP enables AI supply chain agents across geographic locations to synchronize logistics automatically. | SSO Network |
| Session Memory | Temporary retention of contextual information during an AI agent’s session to maintain continuity. | Session memory lets chatbots remember user preferences throughout the conversation. | Auth0 Blog |
| Agent Orchestration | Coordinating multiple agents and tools in sequence to complete complex workflows. | Agent orchestration manages a multi-agent travel booking system, coordinating flight, hotel, and rental car agents. | Dynatrace Blog |
| Autonomy Layer | Design layer that allows AI agents to make independent decisions and act on behalf of users. | Autonomy layer enables smart home AI to regulate temperature without manual user input. | Auth0 Blog |
| Workflow Automation | AI-driven automation of multi-step task processes involving one or more agents. | Automating end-to-end customer onboarding using a sequence of AI validation and communication agents. | Dynatrace Blog |
| Capability Discovery | Mechanism where AI agents advertise and discover available skills among peers. | Agents discover which ones can handle language translation to delegate chat support requests. | Auth0 Blog |
| Meta-Protocol Layer | Higher communication layer allowing negotiation and agreement on communication standards among agents. | Enables agents from different vendors to agree on message formats before interacting. | IBM Think Blog |
| Observability | Tools and techniques that provide visibility into AI agent communication and task execution. | Observability helps debug failures in task handoff between AI agents in customer service workflows. | SSO Network |
| Capability Token | Security token granting an AI agent permission to perform actions or access resources. | Capability tokens secure agent access to sensitive database queries during task execution. | IBM Think Blog |
This set highlights the protocols, infrastructure, and operational concepts underpinning agentic multi-agent AI systems, essential for testers and developers working with advanced AI ecosystems
⁂
Want to learn more❓
Follow Quality With Millan or Share it on LinkedIn
