v0.7.1

• Fix truncated action column
• Optimize English i18n strings

v0.7.0

• decouple ik analyzer of ES
• optimize docs
• add deploy action

v0.6.3

• optimize docs
• fix code block style

v0.6.2

• Add ECS deploy action

v0.6.0

This release focuses on MCP semantic enhancements, an enterprise data warehouse example, frontend table layout fixes, and build optimizations.

Changes & Improvements

• MCP Core & System Tools:
  • Upgraded list_tables to list_tables_and_terms to return both table schemas and business glossary terms in a single call.
  • Standardized tool names, parameter schemas, and prompt descriptions for built-in tools (search_metadata, get_table_info, execute_sql, etc.) to improve LLM accuracy.
  • Enhanced the tool debugging UI to display table schemas and term cards together.
• AdventureWorks DW Example:
  • Added PostgreSQL-compatible schemas and data conversion scripts in examples/adventureworks-dw/.
  • Added documentation covering dimensional modeling, business metrics, and MCP service configuration.
• Frontend & UX:
  • Fixed a layout issue where the table bottom and pagination were clipped in small viewports.
  • Improved DataTableShell: compact height when fewer rows are present, and internal vertical scrolling with sticky headers when data overflows.
  • Simplified landing page feature highlights.
• Build & Documentation:
  • Optimized Dockerfile, .dockerignore, and .gitignore for faster builds and smaller image sizes.
  • Streamlined READMEs and updated OpenAPI specifications.

v0.5.0

DatI v0.5.0 - Initial Release

DatI (Data Intelligence) is an enterprise-grade semantic gateway connecting AI Agents with enterprise databases. It combines semantic modeling, fine-grained access control, and standard Model Context Protocol (MCP) interfaces to enable LLMs to query and manipulate business data accurately and securely.

Core Capabilities

• Data Source Connectivity: Native support for mainstream databases (MySQL, PostgreSQL, ClickHouse, Doris) with a unified two-tier catalog/schema architecture.
• Semantic Modeling: Enriches metadata with business terms, column aliases, enum dictionaries, and sample values, backed by Elasticsearch for semantic retrieval.
• MCP Service Generation: Automatically generates Model Context Protocol (Streamable HTTP) endpoints with prebuilt tools for metadata inspection, SQL execution, and parameterized query templates.
• Security & Access Control: Centrally encrypted credentials, subject-level authorization, and user-context-aware query isolation.

New in v0.5.0

• MCP Usage Statistics: Added invocation tracking and metrics for MCP services to monitor tool call patterns and service usage.
• UX & Table Enhancements:
  • Added "View Details" action for quick schema inspection in Subject tables.
  • Added one-click table name copying with tooltip feedback.
  • Added table filtering by comments and descriptions.
  • Added authorization dialogs to the Subjects page and detail view.
• Deployment & Examples:
  • Added a public online demo instance.
  • Added an out-of-the-box MySQL integration example.
  • Optimized Docker Compose configuration and server deployment documentation.
  • Integrated Spring Boot Actuator for health monitoring.

Quick Links

• Online Demo: http://47.99.122.223:18085/ (Account: demo / demo123)
• Deployment Guide: docs/deployment.md
• Architecture Overview: docs/architecture/overview.md
• Showcase Example: Family Finance Assistant