Nymrio
Intermediate 5 min readingLesson 6 of 6

MCP, REST, A2A, skills and extensions

Interoperability has several layers.

M06 · Concept / external labPractice: 15 minReviewed: Oct 6, 2026

What you will learn

  • Explain: host, client and server
  • Apply the idea in an example: MCP, REST, A2A, skills and extensions
  • Recognize limitations and verify the exercise outcome

Interoperability has several layers. Choose by what needs exchanging: data, tools, instructions or tasks between agents.

This is a conceptual or external lab. It does not assume the application exposes every control described.

How it works, step by step

Host, client and server

MCP is a protocol connecting AI applications to external capabilities. The host provides the experience, the client manages the connection, and the server exposes tools, resources or prompts according to support. Tools execute operations; resources provide addressable content; prompts provide templates. Not every server implements all three. Transport carries messages, authentication identifies callers, and authorization controls access. Check protocol revision and supported extensions at both ends. Compatibility cannot be inferred from an “MCP” label alone.

The model proposes, the tool executes

Tool calling produces a call request with a name and arguments. The application validates the schema, permissions and limits before execution. The result returns to context for the next step. For example, get_product_stock takes a SKU and returns a quantity; its description should explain that it does not reserve products. Distinguish empty results, validation errors, missing access and timeouts. Valid JSON does not prove the arguments are correct or the action is authorized. For external effects, also consider duplicate execution.

A workflow has explicit structure

A node is a stage, an edge is a transition, and state carries results between stages. A workflow has an application-defined path, although the model may choose some branches. Agentic systems include both workflows and dynamically acting agents. Nymrio offers ReAct, Reflection, Sequential, Orchestrator, Router, Parallel and Plan-and-Execute. Each changes coordination; it does not automatically make a model an expert. More agents mean more calls and contracts between roles. Start with the simplest solution that passes your tests.

The visual map

MCP, REST, A2A, skills and extensions Follow the solid arrows for the main flow. Dashed blue arrows supply data or context; dashed pink arrows show feedback or returning results. Colors and shapes distinguish models, stores, decisions and outputs. Connections: AI host application → Choose next action; MCP service → Result to verify; Result to verify → Human verification; REST API → Result to verify; A2A agent communication → Result to verify; Capabilities and constraints → Choose next action; Access permissions → Choose next action; Choose next action → MCP service; Choose next action → REST API; Choose next action → A2A agent communication; Skill instructions → AI host application. M06 · Relationship map MCP, REST, A2A, skills and extensions Processing AI host application Data Capabilities and constraints Decision / control Access permissions Decision / control Choose next action Tool / service MCP service Tool / service REST API Tool / service A2A agent communication Data Skill instructions Output Result to verify Decision / control Human verification Main flow Data and context Feedback and return

Follow the solid arrows for the main flow. Dashed blue arrows supply data or context; dashed pink arrows show feedback or returning results. Colors and shapes distinguish models, stores, decisions and outputs. On smaller screens, scroll horizontally to follow the entire diagram.

A complete example

REST calls a conventional API; MCP discovers and uses AI-facing capabilities; A2A targets agent collaboration. Skills are reusable instructions. MCP Apps, Tasks and Skills over MCP are extensions requiring explicit support.

Try it yourself

  1. Classify stock integration, a reusable guide and a task delegated to another agent.
  2. Record the input, source and expected outcome before running the experiment. Use only the fictional data in the example.
  3. Follow the diagram stages. At every step record what information is received and produced; do not confuse intermediate output with the final outcome.
  4. Repeat after removing necessary information or making the input ambiguous. Check whether the system clarifies, stops or invents an answer.
  5. Compare with the explained solution. Keep the configuration, date, result and an explanation for differences. Change one thing and retest.

An explained solution

Choose REST for a fixed integration, MCP for discoverable tools and A2A where collaboration semantics justify it. A successful exercise lets you show the connection between input, stages and outcome. When information is missing, a cautious answer is more useful than invented details. Compare more than style: check conditions, sources and operations too.

When it helps and what can go wrong

An MCP label does not automatically enable extensions. Choose this approach when it improves a measured need. Keep a simple baseline and compare outcomes using identical inputs. One successful example does not establish reliability in every situation.

Check your understanding

Does a server with tools automatically offer resources and prompts?

No. Capabilities are optional and must be discovered and checked.

Who executes a call proposed by the model?

The application or host service after validation. The model does not automatically gain unrestricted access.

What outcome should this exercise produce?

Choose REST for a fixed integration, MCP for discoverable tools and A2A where collaboration semantics justify it.

Words to remember

  • MCP server: A service exposing capabilities through Model Context Protocol.
  • Tool calling: A structured request to use an external capability.
  • Nod: A stage in an execution graph.

Sources and your next step

To prepare: M05 — Using the application as an MCP server