v0.1.9

Release Notes - sql-agent v0.1.9

Overview

We are pleased to announce the release of sql-agent v0.1.9, introducing a fully implemented CLI interface and an updated licensing model. This version enhances usability and provides clearer legal terms for both open-source and commercial use.

Key Changes

• CLI Implementation

  • Description: The sql-agent-tool command-line interface is now fully functional, supporting interactive database management with PostgreSQL. Key commands include:
    • init: Initialize database schemas.
    • query: Execute natural language queries.
    • shell: Start an interactive SQL shell.
    • update: Manage database updates (disabled in read-only mode by default).
  • Configuration: Loads config.yaml from the current working directory by default, with support for custom paths via the --config option.
  • Impact: Offers a powerful alternative to script-based usage, improving accessibility for users.

• New Licensing Model

  • Description: Transitioned from the MIT license to Apache 2.0 for open-source use, complemented by a dual-licensing approach with a commercial license defined in COMMERCIAL.md.
  • Details:
    • Apache 2.0 provides enhanced patent protection and redistribution terms, suitable for community development.
    • The commercial license (COMMERCIAL.md) supports proprietary applications.
    • Deprecated EULA.md has been removed to streamline licensing documentation.
  • Impact: Ensures robust open-source terms while enabling commercial flexibility.

Installation

To get started with v0.1.9:

pip install sql-agent-tool

Explore available options:

sql-agent-tool --help

Resources

• Source Code: https://github.com/Dadiya-Harsh/sql-agent
• Commercial License: https://github.com/Dadiya-Harsh/sql-tool/blob/main/COMMERCIAL.md

Feedback

We welcome your input on this release. Report issues or suggest improvements via the repository.

---

Full Changelog: v0.1.8...v0.1.9

v0.1.8

updated pyproject.toml file

v0.1.7

Ready for building conda library, now users can download it as conda library too.

Full Changelog: v0.1.6...v0.1.7

v0.1.6

updated required library in pyproject.toml file and added new test cases

v0.1.5

New release installation issue is resolved now..

Full Changelog: v0.1.4...v0.1.5

v0.1.4

📦 SQL Agent Tool v0.1.4

This release updates requiremnts.txt

Developed By - @Dadiya-Harsh, @Dadiya-Harsh-wappnet

🔗 View on PyPI

Full Changelog: v0.1.3...v0.1.4

v0.1.3

📦 SQL Agent Tool v0.1.3

This release introduces an interactive chat interface and enhanced LLM flexibility, empowering users to select their preferred language model for natural language to SQL conversion.

🔧 Updates

• Chat Interface: Added an interactive chat mode for real-time query input and execution.
• LLM Selection: Users can now choose from multiple LLM providers (Groq, Gemini, OpenAI, DeepSeek) to generate SQL, tailored to their needs.

📊 Sample Code

# Launch chat interface and select LLM

   #this are contents of test1.py for testing of tool
   from sql_agent_tool.model import SQLAgentTool, DatabaseConfig, LLMConfig

   config = DatabaseConfig(
      drivername="postgresql",
      username="postgres",
      password="password",
      host="localhost",
      port=5433,
      database="P2"
   )
   llm_config = LLMConfig(provider="gemini", api_key=LLM_API_KEY, model="models/gemini-1.5-flash", max_tokens=500)

   sql_tool = SQLAgentTool(config, llm_config)
   try:
      print("\nQuery 1:")
      q1_start = time.time()
      result = agent_tool.process_natural_language_query("what are top courses purchased by maximum students?")
      print(f"Query 1 total time: {time.time() - q1_start:.2f} seconds")
      if result.success:
         print(f"Query executed successfully, found {result.row_count} results:")
         for row in result.data:
               print(row)

      print("\nQuery 2:")
      q2_start = time.time()
      result2 = agent_tool.process_natural_language_query("Are there any student named harsh?")
      print(f"Query 2 total time: {time.time() - q2_start:.2f} seconds")
      if result2.success:
         print(f"Query executed successfully, found {result2.row_count} results:")
         for row in result2.data:
               print(row)
   except Exception as e:
      print(f"Error processing queries: {e}")
   finally:
      agent_tool.close()
      print(f"Total time: {time.time() - start_time:.2f} seconds")

📊 Sample Output

Selected LLM: Gemini
Enter your query: Show me all courses with the most purchases
Generated SQL:
SELECT c.id, c.title, COUNT(p.id) as purchase_count
FROM courses c
JOIN payments p ON c.id = p.course_id
GROUP BY c.id, c.title
ORDER BY purchase_count DESC
LIMIT 500;
Parameters: {}
Query executed successfully, found 14 results:
{'id': ..., 'title': 'Social Media Marketing (SMM)', 'purchase_count': 6}
...

Developed By - @Dadiya-Harsh, @Dadiya-Harsh-wappnet

🔗 View on PyPI

---

Notes

• Version: Bumped to v0.1.3 since this is a significant feature update (chat interface + LLM choice) beyond the initial v0.1.2.
• Sample Code: Assumed test1.py now supports --chat and --llm flags for the chat interface and LLM selection. If your implementation differs (e.g., no CLI flags), let me know, and I’ll adjust it!
• Sample Output: Adapted from your latest Groq-based run, simplified for brevity.

Full Changelog: v0.1.2...v0.1.3

SQL Agent Tool v0.1.2

📦 SQL Agent Tool v0.1.2

This is the initial release of sql-agent-tool, now available on PyPI!

🔧 Features

• Safe SQL query execution using SQLAlchemy
• Natural language to SQL (via Groq LLM)
• Schema reflection & parameter extraction
• Custom exception handling
• Pytest-based test suite with temporary tables

📊 Sample Code

result = sql_tool.process_natural_language_query("Tell me about user named harsh")

📊 Sample Output

Extracted parameters: {'search_pattern': 'harsh'}
SQL with parameters: 
-- Find user by first name or last name
SELECT *
FROM users
WHERE first_name ILIKE :search_pattern
   OR last_name ILIKE :search_pattern
LIMIT 500;
-- Parameter: search_pattern = '%harsh%'

Query executed successfully, found 1 results:
{'id': 1, 'first_name': 'Harsh', 'last_name': 'Dadiya', 'email': 'harshd.wappnet@outlook.com', ...}

Developed By - @Dadiya-Harsh, @Dadiya-Harsh-wappnet

🔗 View on PyPI

SQL Agent Tool v0.1.1

📦 SQL Agent Tool v0.1.1

This is the initial release of sql-agent-tool, now available on PyPI!

🔧 Features

• Safe SQL query execution using SQLAlchemy
• Natural language to SQL (via Groq LLM)
• Schema reflection & parameter extraction
• Custom exception handling
• Pytest-based test suite with temporary tables

📊 Sample Output

result = sql_tool.process_natural_language_query("Tell me about user named harsh")
Extracted parameters: {'search_pattern': 'harsh'}
SQL with parameters: 
-- Find user by first name or last name
SELECT *
FROM users
WHERE first_name ILIKE :search_pattern
   OR last_name ILIKE :search_pattern
LIMIT 500;
-- Parameter: search_pattern = '%harsh%'

Query executed successfully, found 1 results:
{'id': 1, 'first_name': 'Harsh', 'last_name': 'Dadiya', 'email': 'harshd.wappnet@outlook.com', ...}

Developed By - @Dadiya-Harsh

🔗 [View on PyPI]

Full Changelog: https://github.com/Dadiya-Harsh/sql-tool/commits/v0.1.1