Hey, does anyone have a good example of how to integrate an MCP Server as a Tool for a Sub Agent to use that is part of a Stateful multi agent setup? Thanks!

Hi - @Brandon Hancock. I've been loving your content so far and I've been learning more about the Context Window and timeouts from the Claude Crash Course video. If I have created a Project in Claude (free version) and I give it the Masterplan Prompt for example, it asks me a bunch of clarifying questions; should I be opening a new chat at that point and then answer the questions in a new chat to avoid any timeout issues? In general, Will it remember where the last chat ended or do I need to continue until the request has come to a natural conclusion? I assume the context window in the Claude free version is much smaller than the paid? I have already been timed while it was generating the Masterplan.md file. I do have access to Gemini 3 Pro via my company which should be more generous? (I can still use the same principles as outlined in your Claude Course) Thanks

Hi, I've just joined the Community. I have started my journey into AI Agents! I've been following the ADK Crash Course Video and have downloaded the Git Repo. I have an Idea about developing a Sequential Agent to process DOCX Files and make an API Call to an endpoint. Initially, my first step would be to attach a doc and get the agent to save the document Images to a folder on my machine (MAC) and produce a JSON Object of that Doc ready to be saved into Memory for the next Agent to retrieve. As a newbie to Agents and Python, first - does anyone have any resources or a solution to parsing a DOCX file in a Sub Agent? Thanks for your help....i'm sure there will be more questions! Let me know if you need any further info. Thanks