AI assistants (MCP)
These docs are available to AI coding assistants through a Model Context Protocol server hosted by Scalar. Point your assistant at it and it can search the guides and the Partner API reference while it writes your integration.
| Endpoint | https://mcp.scalar.com/@payroo/payroo-embedded-03e22679 |
| Transport | Streamable HTTP, no authentication |
| Tools | search-documentation (question → the matching pages and sections) |
| Tracks | the current publish of these docs |
Add it to your assistant
Claude Code
claude mcp add --transport http payroo-embedded https://mcp.scalar.com/@payroo/payroo-embedded-03e22679
Cursor, Windsurf, VS Code (Copilot): add to the editor's MCP settings (.cursor/mcp.json, ~/.codeium/windsurf/mcp_config.json, .vscode/mcp.json):
{
"servers": {
"payroo-embedded": {
"type": "http",
"url": "https://mcp.scalar.com/@payroo/payroo-embedded-03e22679"
}
}
}
Cursor and Windsurf use mcpServers as the top-level key instead of servers.
Claude Desktop: Settings → Connectors → Add custom connector, with the endpoint above.
What to ask
- "How do I mint an employer session for a user who is not a Payroo user yet?"
- "Which webhook events exist and how do I verify the signature?"
- "What does
409 payrun_not_approvedmean and what should my server do?"
The assistant reads the same pages you do; the answers quote them. Nothing you ask reaches Payroo's API: the server has no credential and makes no calls.