← sources

New to Claude Code, how do you all use it?

source: reddit (reddit:ClaudeCode)published 2026-11-16 22:13ingested 2026-11-17 02:40

retained text

First of all, I know I can look this stuff up myself, and I am doing my own research. This post is mostly a way to get out all my questions out at once and see how my new peers actually work.

I’m new to Claude Code, and to Claude in general. I just moved over from ChatGPT, where I wasn’t even using Codex. I was just copying and pasting code blocks out of the chat. Since getting Claude last week I’ve been playing with Claude Code and having a good time with it, mostly building Home Assistant dashboards and a few other things. Right now I use either the Claude app or VS Code, depending on what I’m doing. But I feel like I’m not using it to its full potential, and some of the ways I already use it are probably inefficient.

**1. Working across two devices**

Right now I keep one main private GitHub repo. I push and pull between my laptop and desktop, and I use handoff notes for whatever I’m working on. Once I get into bigger projects and start hitting my limits faster, I’m guessing that spending tokens on rereading handoffs isn’t the move. How do you handle this? Is there a better setup for people on multiple devices, or once usage becomes an issue, is the only answer to stay in one session and stop switching back and forth?

**2. Getting Claude Code to deploy where I want**

I have some self-hosted services I’d like Claude to have access to, but I’m also paranoid and don’t want to give it too much control, ya know? For example, my HA dashboard pulls directly from a folder in my repo that Claude pushes to, and that feels like a lot of steps in between.

I know about API tokens because I’ve used a couple to pull data into Home Assistant, but that’s the extent of it. When I asked Claude, it mentioned MCP, SSH keys, and the API tokens. So my main question is: how do you actually connect Claude Code to the services you use it with? whether its for coding the back end, and using it within the service itself for whatever uses there might be.

**3. Agents and harnesses**

I’m writing this before I’ve researched it, and honestly I have no clue what these are. I know what the model is. My guess is that an agent is a bot that uses a specific model as its brain, but I don’t know how I’d actually use one. I keep seeing people talk about “workflows” too, and I’m not sure what those are or how they fit in with agents. This is what I meant about getting the most out of my subscription. I get using it for whatever I need in the moment, but how do I put it to good use even when I’m not actively working with it?

I’ve also seen videos of people building whole “towns” of agents, where each one has its own tasks and role. I’m assuming that’s out of reach on the Pro plan, but I’m curious how it works.

**4. Any other tips?**

If anyone has tips, tricks, or anything else that belongs in a starter guide, I’d really appreciate it. Hopefully this thread helps other people who are just getting started with Claude too.

Whether you’re here for answers, giving answers, or just looking around, thank you. And a special thanks to everyone who read all the way through and to everyone who comments. I appreciate the help as I get started with this.

document fetched 2026-11-17 00:40 · canonical https://reddit.com/r/ClaudeCode/comments/1xy2yvb/new_to_claude_code_how_do_you_all_use_it