Agentic Web Protocols: UCP, AP2, and MCP Standards
As of 2025, AgentFi utilizes open standards to enable AI agents to discover, interact with, and transact across the web. These protocols, including the Universal Commerce Protocol (UCP) and Agent Payments Protocol (AP2), transform traditional websites into machine-readable commerce profiles. By implementing discovery standards like llms.txt and the Model Context Protocol (MCP), the framework ensures agents can execute tasks with verifiable user intent across the entire agentic discovery surface.
What is the Universal Commerce Protocol (UCP)?
The Universal Commerce Protocol (UCP) converts standard websites into discoverable commerce profiles that AI agents can autonomously read and navigate.
* Transforms sites into structured commerce profiles for machine consumption.
* Enables AI agents to identify products and services programmatically.
* Standardizes interaction patterns for agent-led commerce.
* Integrates with Schema.org vocabularies for structured data representation.
Learn more: UCP explained: the Universal Commerce Protocol
How does the Agent Payments Protocol (AP2) work?
The Agent Payments Protocol (AP2) provides a secure framework for AI agents to process financial transactions on behalf of users using cryptographically signed mandates.
* Uses cryptographically signed mandates to authorize agent spending.
* Ensures all agent-initiated payments have verifiable user intent.
* Provides a secure handshake between users, agents, and merchants.
* Standardizes payment requests for seamless agentic checkout experiences.
Learn more: AP2 explained: the Agent Payments Protocol
What are the standards for agentic discovery?
Agentic discovery standards define the specific endpoints and file structures AI agents scan to understand a website's capabilities before taking action.
* llms.txt files provide concise, markdown-based documentation for LLM consumption.
* .well-known endpoints offer standardized paths for agent discovery files.
* XML sitemaps (per the Sitemaps Protocol) provide foundational crawl maps for AI engines.
* These standards ensure a site is visible to 100% of the agentic discovery surface.
Learn more: Agentic discovery standards
How does the Model Context Protocol (MCP) function?
The Model Context Protocol (MCP) establishes a standardized interface that connects Large Language Models (LLMs) to the tools, data, and context necessary for specific tasks.
* Defines a universal interface between LLMs and external tools.
* Facilitates the exchange of real-time data and situational context.
* Reduces friction when integrating third-party APIs into agent workflows.
* Ensures consistent data formatting across different model architectures.
Learn more: MCP explained: the Model Context Protocol