MCP Hands-On Training Project

I want to walk through the entire process of wiring VS Code (with Codex) to an MCP (Model Context Protocol) server and then exposing the data through a simple web dashboard and a chatbot assistant page.

The core of the exercise is the live connection: during the sessions we will spin up an MCP server that can read a small set of local sample project files while also pulling public Steemit data, so I can see both data sources side-by-side. Once that bridge is proven, you will guide me in:

• Configuring VS Code so Codex can hit the MCP endpoint, inspect the model context and run quick prompts
• Building a minimal dashboard that visualises the mixed data (local sample files + Steemit posts) in real time
• Adding a lightweight chat interface where I can ask questions like “show open tasks” or “summarise today’s Steemit stats” and get responses sourced from the MCP server

Please make this a hands-on mentorship. I expect short screen-share sessions, small coding assignments I complete between calls, and clear explanations of each step. By the end I should be able to reproduce the setup from scratch and extend it on my own.

Acceptance criteria
– MCP server started locally, reachable from VS Code and a browser
– Live read of at least one local file type and one Steemit endpoint
– Dashboard page showing two or three example widgets (tables, charts or simple lists)
– Chat page answering at least three sample queries correctly
– All code in a Git repo with concise README covering setup commands, environment variables and troubleshooting tips

Preferred stack: Node.js or Python for the server, simple React/Vite front end (I’m flexible if you have quicker suggestions). Let’s keep the tooling lightweight and focused on demonstrating the protocol, not polishing UI.

If this sounds straightforward, outline your plan and the number of sessions you’d need so we can get started right away.

---
來源：Freelancer.com | 出價次數：19 | 剩餘時間：7 日 | 專案類型：固定預算