Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/jorgeurtubiam-ship-it/Gulin_ia/llms.txt

Use this file to discover all available pages before exploring further.

GuLiN’s DB Maestro feature provides native connections to multiple databases directly inside your terminal. Tell the AI which database you want to work with, hand it the credentials once, and from that point forward you can run queries, generate complex SQL, and render the results as interactive dashboards — all without leaving the terminal window.

​
Supported Databases

Oracle

Native driver via sijms/go-ora/v2. Supports standard Oracle connection strings and TNS aliases.

PostgreSQL

Powered by lib/pq. Works with any PostgreSQL-compatible endpoint including Amazon RDS and Supabase.

MySQL

Uses go-sql-driver/mysql. Compatible with MySQL 5.7+, MariaDB, and PlanetScale.

Microsoft SQL Server

Via microsoft/go-mssqldb. Full support for Azure SQL Database and on-premise SQL Server instances.

MongoDB

Backed by the official go.mongodb.org/mongo-driver. Supports Atlas clusters and self-hosted replica sets.

SQLite

Embedded driver mattn/go-sqlite3 — no server required. Ideal for local development and file-based datasets.

Dremio

Big-data SQL analytics via the API Manager integration. See the Dremio integration page for the full setup guide.

​
Connecting to a Database

Tell the AI which database engine you want to use and provide the connection credentials. GuLiN will establish the connection and keep it available for the rest of the session. 
Connect to my PostgreSQL database at localhost:5432, database "analytics", user "readonly", password "s3cr3t".
Once connected, you can issue any follow-up query in plain language. The AI interprets your intent, translates it to the appropriate query language for the engine, executes it, and returns the results directly in the chat.
Credentials you provide are passed to GuLiN’s secret store, which encrypts them in the local database. They are never stored in plain text or sent to any external service.

​
Generating Queries

Describe what you want in natural language and the AI generates the SQL (or MongoDB aggregation pipeline) for you. You do not need to know the exact table or collection names up front — GuLiN can introspect the schema automatically. 
Show me the total revenue per region for the last 90 days, broken down by product category, sorted descending.
The AI translates that description into a SQL query, runs it against the connected database, and streams the results back.
For complex operations — multi-table joins, window functions, conditional aggregations — describe the business question rather than the SQL shape. For example: “Which customers placed more than five orders last quarter but haven’t ordered anything this quarter?” GuLiN will craft the full query including the necessary joins and date filters.

​
Visualizing Results

Query results can be rendered as interactive dashboards directly in the terminal using recharts-based components. GuLiN supports the following chart types:
TypeKeyBest for
BarbarCategorical comparisons
LinelineTrends over time
AreaareaAccumulated volume
PiepieProportional distributions
RadarradarMulti-variable comparisons
ComposedcomposedCombined bars and lines
GridgridTabular data with search and filter
To render a dashboard, the AI emits a fenced code block containing dashboard metadata. GuLiN’s renderer detects this metadata and replaces the raw block with the interactive chart widget.

​
Dashboard Metadata

Dashboards are configured through three metadata keys embedded in the AI’s response:
KeyDescription
dashboard:typeChart type — one of bar, line, area, pie, radar, composed, grid
dashboard:titleHuman-readable title displayed above the chart
dashboard:dataJSON array of data objects to plot
Example of a complete dashboard metadata block: 
{
  "dashboard:type": "bar",
  "dashboard:title": "Monthly Revenue by Region — Q1 2025",
  "dashboard:data": [
    { "region": "North America", "revenue": 142500 },
    { "region": "Europe",        "revenue": 98200  },
    { "region": "Asia Pacific",  "revenue": 76800  },
    { "region": "Latin America", "revenue": 31400  }
  ]
}
GuLiN captures the first valid data burst and freezes the render state, preventing constant re-renders during long-running queries.

​
Exporting Data

Every rendered dashboard includes built-in export options:
  • Download as PNG — Exports the chart as a high-resolution image with an optimized background, ready for reports or presentations.
  • Export grid data — When a grid chart is active, you can export the full tabular dataset as a CSV file directly from the widget toolbar.

​
Security

All database credentials are handled by GuLiN’s secret store, an encrypted local database that lives entirely on your machine. Credentials are:
  • Encrypted at rest using the local secret store.
  • Never hard-coded into prompts or sent to AI providers.
  • Injected dynamically by the agent at query time.
If you need to update or remove stored credentials, ask the AI to delete the entry from the secret store and register new ones, or manage them directly through the API Manager.

Build docs developers (and LLMs) love

Get started for freeTalk to us