feat(agent):Add excel data manus multi-agent application.​​

[ZhikunZheng]

Description

This PR introduces a complete Excel data manus multi-agent application for DB-GPT, enabling intelligent table conversations and automated data acquisition and enhancement. The application supports end-to-end automation from data acquisition to comprehensive analysis of tabular data.

Key Features Added

1.Excel Table Agent & insert_action:

A specialized agent designed for handling Excel and CSV data with capabilities for single-table and multi-table analysis.

2.WebSearch Agent & websearch_action:

An intelligent agent that evaluates whether existing table data sufficiently covers user query requirements, and performs web searches to supplement relevant data when gaps are identified.

3.Three Agent Examples:

Ready-to-run background examples to facilitate developer understanding and further optimization:

• db_create: Code example demonstrating excel_table_agent functionality;
• auto_search_agent: Code example showcasing WebSearchAgent capabilities;
• data_manus: Comprehensive example utilizing the full excel data manipulation multi-agent system

Comprehensive Documentation:

Detailed user guide including

• Data preparation instructions using SQLite as an example;
• Application creation and usage guidelines

How Has This Been Tested?

The application has been tested with multiple Excel documents, demonstrating the complete workflow from data preparation to analysis result output. The multi-agent system successfully:

• Correctly stores multiple Excel table data into SQLite database
• Performs web searches to supplement information based on identified data gaps
• Executes multi-table queries based on user questions

Snapshots:

Include snapshots for easier review.

default.mp4

Checklist:

• My code follows the style guidelines of this project
• I have already rebased the commits and make the commit message conform to the project standard.
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• Any dependent changes have been merged and published in downstream modules