Thomas Atchapero

We’ll need a frontend, so that the end user can use the automatic contract generator. On entering, frontend should load all the Trello lists, where all tasks have been completed. If there’s a list with incomplete tasks, then it shouldn’t appear in the options for generating a PDF. User should be able to generate a PDF, and preview the result in the browser.

MCP (Model Context Protocol) is an open standard protocol for connecting AI applications to external systems, data sources, tools and workflows. It acts as a universal bridge between AI applications and data sources and tools.

Today the plan is to hook up our backend to an LLM API. We’ll be using the LLM API to generate a contract as a Word/PDF. To do that, we’ll provide the Trello data, along with a contract template for the LLM API to build our contract.

When you install a code agent on your system, you are basically inviting a guest into your house, and trusting that said guest doesn’t take your keys or credit cards, or tells outsiders where your keys are, or what your credit card information is.

To re-iterate on what the next 3 steps are in the current plan, it is: 1. A Java program that can call the Trello API. 2. Extend the Java program to call OpenAI’s API. 3. Test the quality of the generated contracts. And today we’re going to be looking making a Java program that can call the Trello API.

Before we begin building the project, we should have a battle plan: What should we name the project, what’s the core issue, etc. At the end of this article, we’ll have a plan for the next steps, that we can follow up on in the next couple of articles.