What is MCP, and why do legacy systems need it?
Short answer: the Model Context Protocol (MCP) is an open standard that lets AI agents discover and call external tools in a uniform way. Legacy systems need it because they usually expose SOAP endpoints, database schemas or message queues that no agent can use directly.
What MCP does
An MCP server publishes a list of tools, each with a name, a description and a JSON Schema for its inputs. An MCP client (an agent or an AI application) reads that list and can call any tool. The agent does not need custom glue code for each system; it speaks one protocol.
The legacy gap
Many valuable systems predate modern agent tooling:
- SOAP/WSDL services with XML envelopes and verbose contracts.
- Relational databases where the real business logic sits in tables and stored procedures.
- Message queues such as Kafka, RabbitMQ or SQS.
Agents are good at reading tool descriptions and JSON, and poor at hand-assembling SOAP envelopes. Someone has to translate. Writing that translation by hand for every operation is slow and error-prone.
What an MCP wrapper gives you
- One discoverable interface for many back-end systems.
- A single place to enforce validation, permissions and audit logging.
- No changes to the legacy system itself.
Where legacy2mcp fits
legacy2mcp automates the translation. Today it generates an MCP server from a WSDL; database and queue adapters are planned. It is open source under Apache 2.0 on GitHub.
FAQ
Is MCP a replacement for REST?
No. MCP is a layer designed for agents to discover and call tools. It often sits in front of REST, SOAP or database back ends. See MCP vs REST APIs for AI agents.
Do I have to modify my legacy system?
No. The MCP server sits in front of it and translates calls.
Is it safe to let an agent call a legacy system?
It can be, with guardrails. Read MCP server security: safe defaults.