GuanCLI Tool Suite Overview
Overview
This chapter introduces the Guandata BI command-line tool suite: GuanCLI, GuanVis, and GuanETL. All three are designed for AI Agents and command-line environments, but they solve different problems.
- GuanCLI is the general command-line tool for Guandata BI. It connects to BI environments, manages login authentication, queries datasets/pages/cards/ETL and other resources, and supports data preview, task troubleshooting, form operations, and ChatBI calls.
- GuanVis generates visualization resources. Based on GuanCLI login and query capabilities, it lets AI read dataset structures from natural-language requirements and automatically generate and publish BI dashboards.
- GuanETL supports ETL development and debugging. Based on GuanCLI login and query capabilities, it lets AI open, create, modify, check, preview, save, run, and schedule ETL flows.
In short, GuanCLI connects to and queries BI, GuanVis generates dashboards, and GuanETL develops and debugs ETL.
Who Uses These Tools
| User | Typical Tasks |
|---|---|
| Data analyst | Find datasets quickly, review field definitions, preview data, and locate page or card sources. |
| IT / O&M | Check login environments, troubleshoot failed tasks, confirm ETL status, and help locate permission issues. |
| BI developer | Let AI read dataset structure, draft dashboards, modify ETL drafts, and validate results before saving. |
| Enterprise AI Agent | Call BI capabilities in a controlled environment for query, analysis, troubleshooting, and resource generation. |
If business users only want to ask data questions directly in an enterprise AI assistant and the assistant cannot install or call GuanCLI, read GuanMCP User Guide.
Relationship Between the Tools
GuanCLI is the base. GuanVis and GuanETL both depend on GuanCLI.
┌──────────────────────────────────────────────────── ─────────┐
│ GuanCLI (base) │
│ • Connect to Guandata BI environments │
│ • Manage login, tokens, and multi-environment profiles │
│ • Query datasets, pages, cards, ETL, tasks, and metadata │
└───────────────────────┬─────────────────────────────────────┘
│ depend on GuanCLI authentication and query capabilities
┌───────────────┴───────────────┐
▼ ▼
┌───────────────┐ ┌───────────────────┐
│ GuanVis │ │ GuanETL │
│ dashboard │ │ ETL development │
│ generation │ │ save/run/schedule │
└───────────────┘ └───────────────────┘
Before using GuanVis or GuanETL, install and log in to GuanCLI first. GuanVis and GuanETL do not manage BI authentication independently; they reuse the GuanCLI login state.
Which Tool to Use
| Need | Recommended Tool | Description |
|---|---|---|
| Install, log in, switch BI environments, view current account | GuanCLI | GuanCLI handles authentication and environment profiles, and is required before using other tools. |
| Query datasets, pages, cards, ETL, tasks, and other BI resources | GuanCLI | Suitable for resource search, data preview, troubleshooting, and ChatBI calls. |
| Generate BI dashboard drafts from natural language | GuanVis | Suitable for quickly creating pages, charts, and filters based on datasets. |
| Create, modify, check, or save ETL flows | GuanETL | Suitable for data cleaning, field processing, flow debugging, and output validation. |
| Run ETL or configure ETL schedules | GuanETL | Suitable for running saved tasks, waiting for results, or setting scheduled jobs. |
Before You Start
Confirm the following before use:
- Node.js and npm are installed on the current computer.
- GuanCLI is installed and
guancli auth loginhas been completed. - The current login environment is the expected BI environment, and the account has enough permissions for target resources.
- If AI needs to call tools automatically, install GuanVis, GuanETL, and other tools separately and run each tool's
install-skillcommand. - For resource generation, saving, publishing, scheduling, or deletion, prefer a test environment, test directory, or personal directory.
GuanVis and GuanETL can create or modify BI resources. In production, ask AI to explain the target resources, impact scope, and planned operations before saving, publishing, or running.
Difference from GuanMCP
This chapter covers the command-line tool suite for local computers, delivery environments, or any environment where an AI Agent can execute commands.
GuanMCP is for enterprise AI assistants or AI platform integration. It exposes Guandata BI query, metric, metadata, and ChatBI Q&A capabilities as MCP tools so business users can query BI data from Feishu Aily or enterprise-built assistants.
If AI needs to query, troubleshoot, generate dashboards, or modify ETL locally or in a controlled environment, prefer the GuanCLI tool suite. Learn GuanMCP only when the existing AI assistant cannot install or call GuanCLI and must integrate external tools through MCP.
Recommended Reading Order
For first-time users:
- GuanCLI User Guide: install, log in, and learn GuanCLI.
- GuanVis User Guide: generate dashboards with natural language after GuanCLI is available.
- GuanETL User Guide: develop or debug ETL after GuanCLI is available.
If you are not sure whether to choose GuanCLI or GuanMCP, read [[../0-GuanCLI and GuanMCP Comparison|GuanCLI and GuanMCP Comparison]] first.