Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/dataease/SQLBot/llms.txt

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

SQLBot integrates with several popular platforms, either as an MCP tool for AI agents or as an embeddable widget in web applications. This page describes what each integration enables and how to get started.

n8n

Use SQLBot as a tool node in n8n automation workflows.

Dify

Connect SQLBot as an MCP tool inside Dify agent applications.

MaxKB

Call SQLBot from MaxKB agent workflows for live data lookups.

DataEase

Native integration with the DataEase BI platform — SQLBot is built by the DataEase team.

1Panel

Deploy and manage SQLBot directly from the 1Panel application store.
All agent-based integrations (n8n, Dify, MaxKB) connect through the MCP server running on port 8001. Website embedding uses the embedding integration.

​
Platform setup

n8n is a workflow automation platform that can call external tools via HTTP or MCP. You can use SQLBot in n8n to answer data questions as part of a larger automation — for example, generating a sales summary and sending it to Slack, or answering a user’s question in a chatbot workflow.What it enables:
  • Trigger SQL queries from any n8n workflow event (webhook, schedule, form submission, etc.)
  • Combine SQLBot’s query results with other n8n nodes (email, Slack, Google Sheets, etc.)
  • Build AI-powered data pipelines without writing SQL
Setup:
1

Ensure the MCP server is reachable

Confirm that http://<your-sqlbot-host>:8001 is accessible from your n8n instance. If n8n is running in a separate environment, you may need to open port 8001 in your firewall or configure a tunnel.
2

Add an HTTP Request node for authentication

Add an HTTP Request node to your workflow that POSTs to /mcp/mcp_start with your credentials. Store the returned access_token and chat_id in workflow variables using the Set node.
POST http://<your-sqlbot-host>:8001/mcp/mcp_start
Content-Type: application/json

{
  "username": "{{ $vars.SQLBOT_USER }}",
  "password": "{{ $vars.SQLBOT_PASSWORD }}"
}
3

Add an MCP Client Tool node or HTTP Request node

For structured tool use, add an MCP Client Tool node (n8n 1.88+) and point it to http://<your-sqlbot-host>:8001/mcp. For simpler flows, add a second HTTP Request node targeting /mcp/mcp_question directly, passing the token and chat ID from the previous step.
4

Process the result

The mcp_question response includes the generated SQL, query results, and optionally a chart image URL. Wire the output into downstream nodes to send reports, populate spreadsheets, or trigger follow-up actions.
Store SQLBot credentials in n8n’s Credentials or Variables store rather than hardcoding them in node parameters.
Dify is an LLM application development platform. You can register SQLBot as an MCP tool inside a Dify agent, giving the agent the ability to answer questions backed by real database queries.What it enables:
  • Data-grounded answers in Dify chatbot and agent applications
  • SQLBot handles datasource selection and SQL generation; Dify handles the conversation flow
  • Works alongside other Dify tools (web search, knowledge base retrieval, etc.)
Setup:
1

Open the Tools panel in Dify

Navigate to your Dify workspace and go to Tools → Custom → Add Tool.
2

Register the MCP server

Set the Tool Server URL to http://<your-sqlbot-host>:8001/mcp and save. Dify will discover the available operations (mcp_start, mcp_question, mcp_ws_list, mcp_datasource_list) automatically from the MCP schema.
3

Enable the tool in your agent

Open your agent configuration, go to the Tools section, and enable the SQLBot tools you need. At minimum, enable mcp_start and mcp_question.
4

Set agent instructions

In the agent’s system prompt, instruct it to call mcp_start with the SQLBot credentials before answering any data question. Store the token and chat ID in the agent’s context for subsequent calls.
When a user asks a data question, first call mcp_start to obtain an
access_token and chat_id, then call mcp_question with those values.
MaxKB is an open-source knowledge base and agent platform. SQLBot can be added as a tool step in a MaxKB workflow, enabling agents to retrieve live database results in response to user queries.What it enables:
  • Live data lookup alongside MaxKB’s document-based retrieval
  • Natural language questions answered with real query results, not just static knowledge
  • Works in any MaxKB workflow that includes a tool step
Setup:
1

Create or open a workflow

In MaxKB, open the workflow for the agent that needs data access.
2

Add a tool step for mcp_start

Add a Tool step configured to POST to /mcp/mcp_start with your SQLBot credentials. Save the access_token and chat_id from the response into workflow variables.
3

Add a tool step for mcp_question

Add a second Tool step configured to POST to /mcp/mcp_question. Map the user’s input message to the question parameter, and pass the stored token and chat ID from the previous step.
{
  "token": "{{ token }}",
  "chat_id": "{{ chat_id }}",
  "question": "{{ user_input }}",
  "stream": false,
  "return_img": false
}
4

Return the result to the user

Configure the workflow to surface the query result in the agent’s reply. Use return_img: false and stream: false for the cleanest structured output to parse.
SQLBot is developed by the DataEase team — the same team behind the open-source BI and data visualization platform. This shared origin means SQLBot is designed to complement DataEase workflows rather than replace them.What it enables:
  • Use SQLBot for exploratory, conversational analysis while DataEase handles structured dashboards and reports
  • Share the same underlying database connections — configure your datasources once in SQLBot and use them for both ad-hoc queries and DataEase visualizations
  • DataEase users can access SQLBot from a familiar support and release ecosystem
Integration approach:Because SQLBot and DataEase can connect to the same databases, there is no special connector to install. Point both tools at the same database host, configure the datasource in each application, and use them side by side.For organizations running DataEase, the SQLBot embedding integration allows you to add a conversational data assistant directly to any DataEase-adjacent web application.
SQLBot does not read DataEase dashboard definitions or datasets directly. Each tool maintains its own datasource registry. You configure connection details separately in each product.
1Panel is a modern Linux server management panel with an application store. SQLBot is available directly from the 1Panel app store, making it straightforward to deploy and manage without using the command line.What it enables:
  • One-click deployment of SQLBot on any 1Panel-managed server
  • Manage SQLBot’s lifecycle (start, stop, update, uninstall) from the 1Panel UI
  • Automatic port and volume configuration
Setup:
1

Open the 1Panel app store

Log in to your 1Panel instance and navigate to App Store.
2

Search for SQLBot

Search for SQLBot in the app store and click Install.
3

Configure ports and volumes

1Panel will prompt you for port assignments (default 8000 for the web UI and 8001 for the MCP server) and volume paths for persistent data. Accept the defaults or adjust them to fit your server’s layout.
4

Complete installation

Click Confirm. 1Panel pulls the dataease/sqlbot Docker image, creates the container with the configured mounts, and starts the service. Once the health check passes, click the application’s URL to open SQLBot.
After installation, use the 1Panel Upgrade button in the app’s detail page to update SQLBot to newer releases without manually running docker pull.

Build docs developers (and LLMs) love

Get started for freeTalk to us