Integrations
MCP
Seyu ships an MCP server, so AI agents can drive the whole platform: brief in, directed realm out, shipped to a domain. Your agent becomes the studio.
What it is
The Model Context Protocol (MCP) is the open standard that lets AI assistants use external tools. The Seyu MCP server exposes the platform as a set of tools, which means agents like Claude can create realms, give directing notes, and ship the result without a human clicking through the studio.
Setup
Add the server to your client's MCP configuration and set your API key:
mcp.json
{
"mcpServers": {
"seyu": {
"command": "npx",
"args": ["-y", "@seyu/mcp"],
"env": { "SEYU_API_KEY": "sk-seyu-..." }
}
}
}Tools
| Tool | Description |
|---|---|
realm_create | Create a realm from a brief, optionally with a model choice |
realm_note | Apply a directing note to a draft |
realm_status | Fetch state, preview URL, and version history |
realm_ship | Ship a realm to a connected domain |
What that unlocks
Agents can run the whole loop. A few examples of what teams will do with it:
- An agent drafts three realm directions from a product spec and posts the preview links for the team to pick from.
- A launch pipeline ships an updated campaign realm the moment the release is tagged.
- An assistant iterates on directing notes from customer feedback while you sleep.
Pre-season
The MCP server opens with the platform at launch. The tool surface documented here describes the launch scope.