MCP Design

MCP is becoming the USB-C of agent tools. Interviewers ask how you'd design MCP servers for internal APIs, auth, and schema versioning.

MCP is a universal adapter: build one server, and any MCP-compatible client can plug in your tools without custom SDK per IDE.

Scenario

Design MCP server for Jira: tools `create_issue`, `search_issues`; resource `jira://project/KEY`; OAuth to Atlassian; validate project key; return markdown-formatted results for LLM consumption.

Solution

In Interview & System Design, apply MCP Design to this scenario: Design MCP server for Jira: tools `create_issue`, `search_issues`; resource `jira://project/KEY`; OAuth to Atlassian; validate project key; return markdown-formatted results for LLM consumption. Identify the inputs, run the technique, validate the output, and note one thing you would monitor in production.

Do this before moving to the next module — reading alone is not enough.

Open the Code Walkthrough below and run it locally. Change one parameter related to MCP Design (e.g. model, temperature, top_k, or tool name), observe the difference in output, and write 2–3 sentences explaining what changed.

Highlighted lines show where MCP Design happens in the code.

1# MCP Design — minimal example2from openai import OpenAI3
4client = OpenAI()  # create API client5
6# Ask the model to explain this topic7response = client.chat.completions.create(  # core API call for MCP Design8    model="gpt-4o-mini",9    messages=[10        {"role": "system", "content": "You explain mcp design clearly."},11        {"role": "user", "content": f"What is mcp design?"},12    ],13    temperature=0,14)15print(response.choices[0].message.content)  # show output for debugging

• Treating MCP Design as a black box without evaluation
• Ignoring cost and latency in production
• Skipping error handling for mcp design

Quick recap — the most important points from this module.