Claude Code + nyxory
Connect nyxory and Claude Code gets a full operations runtime inside the terminal. Say “deploy this repo” — it builds, ships, and hands you a live URL. Ask “why is it failing?” — it reads the logs and fixes it. One command to start.
The connector opens in the Claude app and you approve it — nothing to paste. Prefer the terminal? The one-line command is right below.
One tap, right in the Claude app. Then ask it to ship.
claude mcp add --transport http --scope user nyxory https://api.nyxory.com/mcp Restart, then tell it: “deploy this repo.”
One product, six services, one project — built by your agent, run by nyxory. Type into the prompt at the bottom; it answers.
› marlo’s ready — deploy it to nyxory
● Marlo is six services — apply form, candidate database, scoring agent, calendar, email, and the recruiter board. Shipping all six into one project.
Live at marlo.nyxory.app — all six healthy.
✶ Shipped in 1m 41s
› now keep it running
● That’s mine now. I watch the whole fleet, restart what stalls, scale to demand, and roll out your agent’s updates as they land. You build it — I run it.
※ recap: one project, six services — built by your agent, run by nyxory.
Illustrative. Your coding agent drives the tools; nyxory ships the project and keeps it running — then hands control straight back.
Deploy straight from the repo you're in
No leaving the terminal. Claude Code clones, picks the right build, ships it, and returns the live URL — from the working directory it is already standing in.
Read live logs and deploy status
When something is off, Claude Code pulls real conditions, history, and streaming logs in the same call, then fixes and redeploys. It debugs like an SRE, not a guesser.
Manage secrets, env, and custom domains
Set environment variables, secret values, and hostnames with automatic TLS — all by asking, all inside the session.
One connect, every project
Add nyxory once at user scope and it is there in every repo you open in Claude Code, with the same tools and the same word.
How do I deploy an app from Claude Code?
Add nyxory as an MCP server with one command — claude mcp add --transport http --scope user nyxory https://api.nyxory.com/mcp — restart Claude Code, then tell it “deploy this repo to nyxory.” It builds the app, ships it, and returns a live URL, all inside your terminal session.
Do I need an API token to connect Claude Code to nyxory?
No. Authentication is OAuth: on first use a browser window asks you to approve, and no token is needed. Tokens exist only for headless or CI setups, created with nyx token create.
What can Claude Code do once nyxory is connected?
Claude Code can deploy apps from a Git repo, stream live logs, read deploy status and history, manage secrets and environment variables, and add custom domains with automatic TLS — all through conversation, without leaving the terminal.
Is nyxory free to start?
Yes. Starting is free: connect, create a project, and deploy your first app. Plans add more resources as your work grows — see nyxory.com/pricing.
Where does nyxory run my apps?
On EU infrastructure. Every project gets isolated resources, and destructive operations are clearly annotated so Claude Code asks before anything irreversible happens.
Add nyxory to Claude Code, or start in the console — free to begin, EU hosted.
Need a hand? Join our Discord or email hello@nyxory.com.