An AI agent that actually operates your servers over SSH — runs commands, edits files, deploys, and monitors — with the guardrails you set.
Organize everything into Projects — each with its own instructions, model, and history. Pick Claude, ChatGPT, or Gemini per project. Attach images, PDFs, and files. It's a proper AI workspace, not a toy chat.
Give a project SSH access and the AI can read logs, grep configs, patch files, restart services, and deploy. You choose the autonomy. Writes are blocked until you enable them — every command is yours to approve.
Pick the AI that fits the job and switch anytime — no lock-in. Give a tricky refactor to Claude, a quick script to ChatGPT, or a huge log to Gemini. Set a default per project, then swap models mid-conversation with one tap. Same tools, same guardrails — you just decide who does the thinking.
Set a default — swap mid-conversation anytime
Dedicated read_file / write_file tools over SFTP mean exact reads and clean writes — no shell-quoting surprises. It reads a file, edits it, verifies it, and moves on.
Not a code snippet to run yourself — an actual, downloadable file. PDFs, Word, Excel, PowerPoint, SVG, Illustrator and Photoshop, plus Markdown, CSV, JSON, HTML and code. Preview it right in the chat, then save to Files or share it anywhere.
Watch your sites and servers with uptime %, response-time graphs, incident history, and a "last checked" clock. Get a push notification the moment something goes down. Export any report to PDF or CSV.
Home Screen and Lock Screen widgets for live server status and your daily activity — tap through straight to the detail. A configurable summary widget shows exactly what you want.
A built-in dashboard: what you accomplished today, focused time, streaks, 7-day activity, and AI spend per model. Your whole workflow, measured.
Projects, chat history, monitors, and settings sync end-to-end through your own iCloud — never our servers. Start a deploy on your iPhone, pick it up on your Mac, and check status on iPad. API and SSH keys stay in the on-device Keychain and are never uploaded.
A native Apple Watch app and complications keep server status one glance away. Feel a haptic tap the instant a site goes down, then Approve & run the AI's fix straight from your wrist — no phone needed.
Pick your AI and paste your server's SSH key — it goes straight into the Keychain.
Plan, Ask, or Auto. Writes stay blocked until you say otherwise. You're always in control.
"Why is nginx 502-ing?" — and watch it investigate and fix, command by command.
Start with a 7-day free trial. Monthly or annual after that — no seats, no metered anything. Model usage is billed by your AI provider — DockPilot is just the deck.
Billed $39.99 once a year
Free for 7 days, then $4.99/mo or $39.99/yr. Cancel anytime.
Fixed a prod incident from a beach in Lisbon. The approval flow is what makes it actually safe to hand over the keys.
First mobile server tool that doesn't feel like typing through a straw. I say "why is this slow" and it just… finds it.
The monitoring widgets alone are worth it. My lock screen tells me everything's green before I'm out of bed.
In the Apple Keychain, on your device. They're never uploaded to us, never sent to any model, and never leave the device except to authenticate directly with your provider or server.
Only what you allow. Four autonomy levels — Plan, Ask, Auto read-only, Full auto — and writes are blocked until you explicitly enable them. Every command is logged.
Only the command output you're actively working with — the exact text a command returns — is sent to the model you chose, to reason about the task. DockPilot never bulk-uploads your filesystem, databases, or logs, and keeps no copy on its own servers.
Claude, ChatGPT, and Gemini — per project, with your own API keys. You pay your provider directly; DockPilot adds no markup and takes no cut.
Anything you can reach over SSH — a VPS, a bare-metal box, a Raspberry Pi, or a cloud instance on AWS, DigitalOcean, Hetzner, Linode, and the rest. Password or key auth, custom ports, and jump hosts are all supported.
Add a project, paste your host, user, and key — or generate a fresh keypair on-device — pick a model, and you're in. It takes about a minute, and there's no agent to install on the server side.
Yes. In Plan and Ask modes, DockPilot shows you the exact command before it touches your server and waits for your tap. You can edit or reject anything, and every action stays in a full, searchable history.
Your projects, command history, and monitoring live on your device and sync privately through your own iCloud — not our servers. We can't read them, and there's no DockPilot account or cloud login to create.
Nothing breaks. The SSH session simply pauses, and DockPilot reconnects and picks up where you left off when you're back online. Long-running jobs keep going on the server regardless of your connection.
Absolutely. DockPilot is also a straight-up SSH client and dashboard — check CPU, memory, disk, processes, and logs at a glance, or drop into a raw terminal. The AI is there when you want it, not in the way when you don't.
No. One purchase, every feature, free updates, Family Sharing included. The only ongoing cost is whatever you pay your own AI provider for usage.
Yes — one universal purchase covers iPhone, iPad, Mac, and Apple Watch. Projects, history, and monitoring sync over iCloud, so you can start on your phone and finish at your desk.
The AI command deck that fits in your pocket.
