It is a remote server — nothing to install. Add it to your MCP client and sign in.
OAuth-capable clients will ask you to sign in and pick an organization. Clients that authenticate with a header need a token with the publish:write scope, sent as an Authorization: Bearer header.
Your IDE agent becomes the CMS client
The agent you already use for code — or the chat window you already have open — can create a site, write its pages and publish it. No separate admin UI in the loop, and no copy-paste between the editor and the browser.
Content that lives with the repository
Because publishing is an API call, documentation and marketing pages can be updated in the same change that alters the code they describe — rather than drifting until someone notices.
Incremental by design
The changes endpoint lets an agent see what moved since last time, so a scheduled run updates what needs updating instead of rewriting the site every pass.
Three steps to a published page
01
Add the server
One command in Claude Code, or paste the connector URL into your client.
02
Authenticate
Sign in over OAuth, or use an API token with the publish:write scope.
03
Ask for a page
The agent creates the site, writes the content and publishes it.
Any MCP client that supports remote streamable HTTP servers. ChatGPT, Claude Code, Claude.ai and Claude Desktop, Cursor and VS Code are the ones Aveiro tests against.
Do I need to install a package?
No. The server is remote — you point your client at a URL rather than installing anything locally.
How does authentication work?
Clients that support OAuth will prompt you to sign in and choose an organization. For clients that authenticate with a header, create an API token with the publish:write scope and send it as a bearer token.
Can an agent publish without me seeing it first?
It can, if you give it a token with publish scope — that is the point for automated runs. Scope the token to what you actually want the agent to do.