Formio Uag MCP server
{
"mcpServers": {
"formio-uag": {
"command": "npx",
"args": [
"-y",
"@formio/uag@latest"
],
"env": {
"BASE_URL": "https://...",
"PROJECT_KEY": "<project_key>",
"ADMIN_KEY": "<admin_key>",
"PORTAL_SECRET": "8080"
}
}
}
}{
"mcpServers": {
"formio-uag": {
"command": "npx",
"args": [
"-y",
"@formio/uag@latest"
],
"env": {
"BASE_URL": "https://...",
"PROJECT_KEY": "<project_key>",
"ADMIN_KEY": "<admin_key>",
"PORTAL_SECRET": "8080"
}
}
}
}{
"mcpServers": {
"formio-uag": {
"command": "npx",
"args": [
"-y",
"@formio/uag@latest"
],
"env": {
"BASE_URL": "https://...",
"PROJECT_KEY": "<project_key>",
"ADMIN_KEY": "<admin_key>",
"PORTAL_SECRET": "8080"
}
}
}
}{
"servers": {
"formio-uag": {
"command": "npx",
"args": [
"-y",
"@formio/uag@latest"
],
"env": {
"BASE_URL": "${input:base_url}",
"PROJECT_KEY": "${input:project_key}",
"ADMIN_KEY": "${input:admin_key}",
"PORTAL_SECRET": "${input:portal_secret}"
}
}
},
"inputs": [
{
"id": "base_url",
"type": "promptString",
"description": "Base url",
"password": true
},
{
"id": "project_key",
"type": "promptString",
"description": "Project key",
"password": true
},
{
"id": "admin_key",
"type": "promptString",
"description": "Admin key",
"password": true
},
{
"id": "portal_secret",
"type": "promptString",
"description": "Portal secret",
"password": true
}
]
}{
"mcpServers": {
"formio-uag": {
"command": "npx",
"args": [
"-y",
"@formio/uag@latest"
],
"env": {
"BASE_URL": "https://...",
"PROJECT_KEY": "<project_key>",
"ADMIN_KEY": "<admin_key>",
"PORTAL_SECRET": "8080"
}
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsoninstall npx -y @formio/uagversion 1.13.2
What it does
The Formio UAG is an MCP server that transforms Form.io JSON schemas into AI-readable markdown, enabling AI agents to understand and work with forms. It supports two main patterns: converting natural language input into structured form submissions, and running autonomous agentic workflows that read and populate form data.
- Convert natural language conversational input into structured data that matches a Form.io schema and submit it as a form entry.
- Build an autonomous workflow where an AI agent reads existing form submission data, applies criteria or rubrics, and writes analysis results back into the form.
- Create a chat-based form interface where all questions and field definitions come from a Form.io form, with no hardcoded field names in your application code.
Tools this server exposes
| Tool | Does |
|---|---|
| form_render | Transforms a Form.io JSON schema into AI-readable markdown format so agents can understand the form structure and requirements. |
| form_submit | Submits structured data as a form entry to a Form.io form. |
| form_context | Injects dynamic context from Form.io into an AI processing cycle to modify agent behavior without retraining. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.