# Description Descriptions let you explain what a table, collection, column, or field means in your actual product. DataLoam includes that context in the Database Index so AI answers can use your language and your domain rules. What to describe [#what-to-describe] * What a table or collection represents. * When a field should be used. * Business meaning that is not obvious from the name. * Important constraints, edge cases, or naming conventions. Example [#example] Instead of only indexing a column named `status`, add a description such as: ```text Current fulfillment state of the order. Use "processing" for paid orders that have not shipped yet. ``` That gives DataLoam more useful context when it generates queries or explains results. # Enum Enum detection helps DataLoam identify fields that behave like enums, even when the database does not define them as formal enum types. What counts as enum-like [#what-counts-as-enum-like] A field is enum-like when it usually contains a small set of repeated values. Common examples include: * `status` * `payment_status` * `role` * `priority` * `type` Why it matters [#why-it-matters] When DataLoam understands enum-like fields, AI can generate more accurate filters and explanations. For example, it can prefer known values such as `shipped` or `refunded` instead of guessing a value that does not exist. Local context [#local-context] Enum-like values are part of the Database Index context DataLoam uses to keep query help grounded in your real schema. # Database Index Database Index is the local context layer DataLoam builds from your database schema, relationships, documentation, and learned values. It helps AI features answer with real database context instead of generic guesses. What it stores [#what-it-stores] * Database structure, including schemas, tables, collections, columns, and fields. * Relationships between entities, such as foreign keys and joined data paths. * Documentation you write for tables, columns, collections, and fields. * Enum-like values DataLoam learns from your database shape. Why it matters [#why-it-matters] AI is only useful when it understands the database it is working with. Database Index gives DataLoam the context needed to generate better queries, avoid invalid fields, and explain database-specific workflows more accurately. Local by default [#local-by-default] The index is built for your local editor workflow. DataLoam uses it to provide schema-aware autocomplete, query assistance, and AI guidance without forcing you to repeatedly describe your database. # Values Values are examples DataLoam can learn from your database shape. They help the Database Index understand what is valid for fields where the type alone is not enough. Why values help [#why-values-help] Many useful filters depend on exact values: * Order statuses such as `processing`, `shipped`, and `cancelled`. * Payment states such as `paid`, `pending`, and `refunded`. * User roles such as `admin`, `manager`, and `viewer`. When DataLoam knows these values, AI can avoid inventing invalid filters. How to use them [#how-to-use-them] Use values as supporting context for fields that are often filtered, grouped, or explained in queries. # Getting Started Start here if you are setting up DataLoam for the first time. Install DataLoam from the VS Code Marketplace, with notes for Cursor users. Connect a database and begin working with DataLoam. # Install DataLoam is distributed as a VS Code extension. You can install it from the VS Code Marketplace, then use the same extension in VS Code-compatible editors such as Cursor. VS Code [#vs-code] Open the marketplace page [#open-the-marketplace-page] Open the DataLoam extension page in the VS Code Marketplace: [Download DataLoam for VS Code](https://marketplace.visualstudio.com/items?itemName=VPALGOsro.dataloam) Install the extension [#install-the-extension] Click **Install** on the marketplace page. Your browser may ask to open VS Code. Accept the prompt, then confirm the installation inside VS Code. Open DataLoam [#open-dataloam] After installation, open the Command Palette with `Ctrl+Shift+P` on Windows or Linux, or `Cmd+Shift+P` on macOS. Search for `DataLoam`, then open the DataLoam view and add your first database connection. Cursor [#cursor] Cursor supports VS Code extensions, so you can install DataLoam from the Extensions view. Open Extensions [#open-extensions] Open Cursor and go to the Extensions view from the activity bar, or open the Command Palette and run `Extensions: Install Extensions`. Search for DataLoam [#search-for-dataloam] Search for `DataLoam` and install the extension published by DataLoam. Use the VS Code Marketplace link if needed [#use-the-vs-code-marketplace-link-if-needed] If DataLoam does not appear in Cursor search, open the VS Code Marketplace page: [Download DataLoam from the Marketplace](https://marketplace.visualstudio.com/items?itemName=VPALGOsro.dataloam) Cursor can also install VS Code-compatible extensions through a `.vsix` file if your setup requires manual installation. After install [#after-install] Once DataLoam is installed, connect a database, let the extension index your schema locally, and start writing queries with schema-aware AI assistance. # Quick Start After installing DataLoam, connect a database and open your first query file. Open DataLoam [#open-dataloam] Open the Command Palette with `Ctrl+Shift+P` on Windows or Linux, or `Cmd+Shift+P` on macOS. Search for `DataLoam`, then open the DataLoam view. Add a connection [#add-a-connection] Create a database connection from the DataLoam view. Choose your provider, enter your connection details, and save the connection for reuse. Let DataLoam index your schema [#let-dataloam-index-your-schema] Open the connection and let DataLoam read the database shape locally. The index helps AI understand tables, columns, relationships, and useful context before it generates queries. Start querying [#start-querying] Create or open a query file, ask DataLoam for the result you need, review the generated query, and run it against your database. # Guides Guides focus on specific workflows and habits that make DataLoam faster to use inside your editor. Move through DataLoam, open connections, and work with queries from the keyboard. # Keyboard Navigation DataLoam is designed for editor-first workflows, so common database actions can be reached without switching to the mouse. Use keyboard navigation when you want to open a connection, move through database objects, run a query, or return to the editor quickly. Open DataLoam commands [#open-dataloam-commands] Use the Command Palette to find DataLoam actions: * Windows and Linux: `Ctrl+Shift+P` * macOS: `Cmd+Shift+P` Search for `DataLoam` to see available commands. Common workflow [#common-workflow] Open a database [#open-a-database] Run the DataLoam command for opening a database connection. Choose the saved connection you want to work with. Move through the database tree [#move-through-the-database-tree] Use the arrow keys to move through schemas, tables, collections, and fields. Expand or collapse the selected item with `Enter`. Open a query file [#open-a-query-file] Select a query file or create a new one from the DataLoam commands. The editor keeps focus on the query surface so you can start writing immediately. Run and inspect results [#run-and-inspect-results] Run the current query from the keyboard, then move between the editor and result panel as needed. Tips [#tips] * Use the Command Palette when you are not sure where a DataLoam action lives. * Keep saved connections named clearly so they are easy to find from keyboard search. * Pair keyboard navigation with Vim mode if you prefer modal editing.