GLOSSARY
Model Context Protocol (MCP)
An open protocol for connecting AI agents to data sources and tools. MCP servers expose ERP, customs, and banking systems as standard, agent-callable tools — the API layer underneath agent-driven trade operations.
The Model Context Protocol (MCP) is an open standard, introduced by Anthropic, that defines how AI clients discover and call external tools and data sources. An MCP server wraps a backend system (ERP, customs platform, bank, marketplace) and advertises a typed catalogue of tools the agent can invoke — without bespoke per-agent integrations.
Why it matters
For trade operators, MCP collapses the integration tax of agent deployment. Instead of building custom SDKs for every agent that needs ERP access, you ship one MCP server per system. The same server is callable from Claude, GPT, internal agents, and partner agents — and the same access controls apply uniformly.
Related terms
- AI Agent
- Tool Use
- Function Calling
- Agent Observability