Get ready for your assistant setup

You are helping me get ready for a scheduled setup call where a specialist (Cody Mitchell) will install a personal always-on AI assistant on a small cloud server that I own. Your job is to walk me through the preparation, ONE STEP AT A TIME. Rules for you: - Give me one instruction at a time, describing what I will actually see on screen. Wait for me to confirm before the next step. - When a step has a choice, tell me which option to pick and why in one line. - I may not be technical. No jargon without a plain explanation. - NEVER ask me to send you, or anyone, a password, API key, or private key. If I try to paste one to you, stop me. Secrets get typed by me, directly into the systems that need them, during the install call itself. Nothing in this prep requires sharing a secret with anyone. Walk me through these stages in order: STAGE 1 - Claude subscription. Check I have a Claude Pro or Max subscription at claude.ai that I can log into. My assistant will run on this plan. If I do not have one, walk me through subscribing (Pro is enough to start). STAGE 2 - The server. Walk me through creating a small cloud server on DigitalOcean (or an equivalent I already use): the cheapest droplet with 1GB or more of RAM, Ubuntu LTS, in a region near me. During creation, when it asks for an authentication method, choose SSH KEY and walk me through creating one on my computer if I do not have one (ssh-keygen with the default suggestions is fine). Make sure I end up with: the server running, its IP address written down, and me able to connect to it (walk me through one test connection from my computer's terminal). STAGE 3 - The specialist's access. So the specialist can install and maintain my assistant, their PUBLIC key needs to be on the server. This is safe: a public key lets them log in with their own private key, and shares no secret of mine. Walk me through adding this exact line to /root/.ssh/authorized_keys on the server (or via my provider's console): ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICjmjD1rdiFskphlCfn+tAgvhBEdSSInGmNOTcjn9JDp cody-operator-codaclaude STAGE 4 - Telegram and my bot. Check I have Telegram installed and signed in on my phone. Then walk me through creating my assistant's bot so it is ready before the call: in Telegram, search for @BotFather (the verified one), send /newbot, pick a display name (my assistant's name), then a unique username ending in "bot". BotFather replies with an HTTP API token. THAT TOKEN IS A SECRET: do not paste it to you or into my report. Walk me through saving it somewhere safe I can retrieve on the call (my password manager, or a private note). The only thing that goes in my report is the bot's @username. Do not send the bot any messages yet; that happens on the call. STAGE 5 - Thinking ahead (no wrong answers, five minutes). Ask me these two questions and help me sharpen my answers into a few clear sentences: a) What do I want this assistant to handle first? (examples: a daily morning brief, research, drafting, reminders, keeping tabs on a project or inbox) b) What should it NEVER do without asking me first? (it already ships with hard blocks on spending money, deleting things, and touching credentials; this is about MY additions) STAGE 6 - The readiness report. When every stage is done, produce a short report for me to send to Cody, containing ONLY: - Confirmation of each stage (done / not done) - My bot's @username (never the token) - The server's IP address and provider - Which region and size I picked - My answers to the two questions in stage 5 - Anything I got stuck on Remind me one final time: this report must contain no passwords, no keys, no tokens. If it does, remove them before sending.