github.devotd.
NewAI

eosphoros-ai/db-gpt v0.8.2 — A data agent that writes SQL and code in natural language

eosphoros-ai/db-gpt is a Python-based, MIT-licensed open-source data agent framework that connects to databases, CSV/Excel files and knowledge bases, writes SQL and code in response to natural-language questions, and generates charts and HTML reports; the latest release is v0.8.2.

Repositoryeosphoros-ai/db-gpt ↗

eosphoros-ai/db-gpt is a Python-based, MIT-licensed open-source data agent framework that connects to databases, CSV/Excel files and knowledge bases, writes SQL and code in response to natural-language questions, and generates charts and HTML reports; the latest release is v0.8.2.

GitHub description: open-source agentic AI data assistant for the next generation of AI + Data products.

stars
★ 20.2K
Forks
3K
Language
Python
License
MIT
Latest release
v0.8.2
Read original source ↗View on GitHub ↗

eosphoros-ai/db-gpt is an open-source agentic AI data assistant that writes SQL and code in response to natural-language questions in order to analyze data. It is written in Python and released under the MIT license, and the repository's latest release is v0.8.2. The list of changes included in this version is not given in the provided material.

Its connection targets are broad, covering not only databases and data warehouses but also CSV/Excel files and knowledge bases, and generating charts, dashboards and HTML reports is the central feature of this project. It is a framework focused on supporting professional data work through SQL analysis and data agent workflows.

The README explains that it supports reusable skills and sandbox execution. A skill is a unit that bundles together a repeatedly used procedure, and sandbox execution refers to running the code you create in an isolated environment. Combining these with retrieval-augmented generation (RAG) and multimodel support, it also presents a path to building AI-native data agents and workflows and applications.

In the score records, only momentum (모멘텀) scores 3.4 points (out of 10), and the freshness, technology and quality scores have no values given. The list of supported models and the isolation level of the sandbox are also absent from this material. The repository was created on April 13, 2023 and has gathered 20172 stars, but until the improvements of v0.8.2 are confirmed, it is difficult to cite it as grounds for introducing new features or replacing existing tools.