An MCP server that hands Cursor, Claude Code and v0 your real components, tokens and patterns, so nothing is guessed.
Build Relay, a dispatch console for a city courier service: a live map, orders, couriers, zones, insights and alerts. Use Orvain.
42 blocks, 61 specs and your tokens, from Orvain.
Six tools behind one connection, wired to the same components, tokens and patterns your engineers ship.
A hosted Model Context Protocol server. Agents call get_component, get_tokens and get_pattern against the live library.
Full prop tables, slots and variants in the model's context, so it knows the API before it writes a line.
Color, radius, spacing and type tokens from one config. Re-theme once and every generation follows.
Figma components mapped to their code, so hand-off and AI generation read the same source of truth.
Describe a screen in plain language and get a composed page built from your components: billing, dashboards, settings.
Every call validated against the schema. Invented props, missing variants and made-up imports are rejected.
The Orvain MCP server works in anything that speaks the Model Context Protocol, with one entry in your editor's config.
One entry in .cursor/mcp.json.
Native MCPOne claude mcp add command.
Native MCPOne entry in Cascade's MCP config.
Native MCPOne entry in .vscode/mcp.json, in agent mode.
Native MCPA drop-in adapter, no setup of your own.
Via adapterOne server URL in its MCP settings.
Native MCPWhat is an MCP server, exactly?
Model Context Protocol is an open standard that lets AI tools call structured “tools” over a connection. Our hosted endpoint exposes your component library (its props, tokens and patterns) so any MCP-aware assistant can query the real source instead of guessing from training data.
Which AI tools are supported?
Anything that speaks MCP (Cursor, Claude Code, Windsurf and GitHub Copilot in VS Code) plus a drop-in adapter for v0. Connecting is a single config entry pointing at mcp.orvain.dev.
How do you guarantee “0 hallucinated props”?
Every generation is validated against the live component schemas. Props that don’t exist, missing required variants, and invented imports are rejected before code is written: the agent is forced to correct itself against the real API.
Does this work with my own theme?
Yes. White-label tokens resolve from one config file. Re-theme once and every subsequent generation (colors, radii, spacing, type) follows your brand automatically.
Join the Orvain Dispatch and get production-ready components, free to keep: React and Vue code plus the Figma frame, yours to ship.
An MCP server that hands Cursor, Claude Code and v0 your real components, tokens and patterns, so nothing is guessed.
Build Relay, a dispatch console for a city courier service: a live map, orders, couriers, zones, insights and alerts. Use Orvain.
Six tools behind one connection, wired to the same components, tokens and patterns your engineers ship.
A hosted Model Context Protocol server. Agents call get_component, get_tokens and get_pattern against the live library.
Full prop tables, slots and variants in the model's context, so it knows the API before it writes a line.
Color, radius, spacing and type tokens from one config. Re-theme once and every generation follows.
Figma components mapped to their code, so hand-off and AI generation read the same source of truth.
Describe a screen in plain language and get a composed page built from your components: billing, dashboards, settings.
Every call validated against the schema. Invented props, missing variants and made-up imports are rejected.
The Orvain MCP server works in anything that speaks the Model Context Protocol, with one entry in your editor's config.
What is an MCP server, exactly?
Model Context Protocol is an open standard that lets AI tools call structured “tools” over a connection. Our hosted endpoint exposes your component library (its props, tokens and patterns) so any MCP-aware assistant can query the real source instead of guessing from training data.
Which AI tools are supported?
Anything that speaks MCP (Cursor, Claude Code, Windsurf and GitHub Copilot in VS Code) plus a drop-in adapter for v0. Connecting is a single config entry pointing at mcp.orvain.dev.
How do you guarantee “0 hallucinated props”?
Every generation is validated against the live component schemas. Props that don’t exist, missing required variants, and invented imports are rejected before code is written: the agent is forced to correct itself against the real API.
Does this work with my own theme?
Yes. White-label tokens resolve from one config file. Re-theme once and every subsequent generation (colors, radii, spacing, type) follows your brand automatically.
Join the Orvain Dispatch and get production-ready components, free to keep: React and Vue code plus the Figma frame, yours to ship.
Orvain MCP