The project is a frontend part of the DIAL Mind Map Studio.
Mind Map enables users to access information through an interactive knowledge graph and natural language. The application pulls data from various sources, including documents, URLs, and other data inputs and then presents it on UI as a interactive knowledge graph, facilitating intuitive and engaging user interactions with information.
Click to watch the demo video
Read more about DIAL Mind Map Studio
Execute npm run dev to run a frontend server in dev mode.
The Mind Map Studio application uses environment variables to configure authentication, API connections, and theming settings. Below is a list of environment variables used in this project.
| Variable | Required | Description |
|---|---|---|
NEXTAUTH_SECRET |
Yes | Secret key for NextAuth.js authentication. Possible to generate with openssl rand -base64 32. |
NEXTAUTH_URL |
Yes | Base URL of the application used for authentication callbacks. |
AUTH_KEYCLOAK_CLIENT_ID |
Yes | Client ID for Keycloak authentication. |
AUTH_KEYCLOAK_SECRET |
Yes | Secret key for Keycloak authentication. |
AUTH_KEYCLOAK_HOST |
Yes | Keycloak server URL. |
AUTH_KEYCLOAK_REALM |
Yes | Keycloak realm name used for authentication. |
THEMES_CONFIG_HOST |
No | The host URL for custom themes configuration. |
THEMES_CONFIG |
No | JSON string pointing to the theme configuration for chat customization. Example. |
DIAL_API_KEY |
No | API key used to authenticate requests to the AI DIAL API. |
DIAL_API_HOST |
Yes | The API host for AI DIAL integration. |
DIAL_CHAT_HOST |
Yes | The host URL for AI DIAL chat integration. |
DIAL_IFRAME_ALLOWED_HOSTS |
No | Comma-separated list of all DIAL application hosts that are allowed to be embedded as iframes (e.g. https://chat.dial.ai,https://builder.dial.ai,https://mindmap.dial.ai). If not set, the app will fall back to DIAL_CHAT_HOST for backward compatibility. |
MINDMAP_IFRAME_TITLE |
Yes | Custom title to display in the iframe when embedding the mindmap application. Must match dial:applicationTypeDisplayName in DIAL schema. |
BUILDER_ALLOW_API_KEY_AUTH |
No | Builder pages only. If set to true, allows authentication via API key on the builder UI without requiring user login. Acceptable values: true, false. Default: false. |
CHAT_DISCLAIMER |
No | A text string which, if set, will be displayed below the chat input. |
GCP_PROJECT_ID |
No | The Google Cloud Project ID where your reCAPTCHA Enterprise configuration is hosted. This value is required by the backend to create and assess reCAPTCHA Enterprise assessments using the official RecaptchaEnterpriseServiceClient. Make sure it matches the project that contains your reCAPTCHA Enterprise keys and security policies. |
RECAPTCHA_SITE_KEY |
No | Site key for Google reCAPTCHA. |
RECAPTCHA_REQUEST_QUOTA |
No | The number of requests a user can make before triggering the reCAPTCHA challenge. |
ANONYM_SESSION_SECRET_KEY |
No | The secret key used for anonymous session cookies when DIAL_API_KEY is set. This variable becomes mandatory when DIAL_API_KEY is configured. It is used to encrypt the anonymous session cookie, which holds information necessary for performing reCAPTCHA verifications. Generate it using openssl rand -base64 32. |
GOOGLE_FONTS_API_KEY |
No | A unique API key required to authenticate and enable access to the Google Fonts API. Without it, font family validation cannot be performed. |
AUTH_UI_MODE |
No | Controls the authentication UI mode: set to popup to use a popup window, tab to open a new browser tab, or sameWindow to authorize in the same window where the mindmap is embedded. Defaults to popup if unset |
DEFAULT_CHAT_MODEL |
No | Default chat model id/slug. If unset, a code-level fallback is used — DEFAULT_CHAT_MODEL. |
AVAILABLE_LITE_MODE_MODELS |
No | A comma-separated list of model identifiers available for the Advanced mode graph generation (e.g., gpt-4.1-2025-04-14,gpt-5-2025-08-07,gemini-2.5-pro). If unset or empty, all models will be available. |
AVAILABLE_CHAT_MODELS |
No | A comma-separated list of model identifiers available for the Chat RAG. If unset or empty, all models will be available. |
LITE_MODE_TOKENS_LIMIT |
No | The maximum allowed token usage across all sources when generating a graph. If the combined token usage of all sources exceeds this limit, graph generation will be blocked, and the user will be prompted to adjust their sources or reduce token usage. |
Important
Make sure to define all required variables in your .env file or provide them through your deployment configuration.