TablePlus AI Assistant
AI agent · Data & analytics agent · Data & analytics by TablePlus
A built-in AI assistant for TablePlus that writes, explains, and runs database queries.
TablePlus AI Assistant is a built-in coding and data agent designed for developers and database administrators using the TablePlus GUI. It assists users by writing, explaining, and executing SQL queries directly within the database workspace. The agent can autonomously query the database to gather context, though it requires explicit user approval for any non-read-only operations. It supports a wide range of backends, including cloud-based LLMs like OpenAI and Anthropic, local models via Ollama, and third-party agents through the Agent Client Protocol (ACP). By integrating directly into the ri
- Languages
- English
- Deployment
- In-app sidebar (macOS)
- Supported Protocols
- Agent Client Protocol (ACP), Model Context Protocol (MCP)
- Local Execution
- Ollama, Codex CLI
Overview
TablePlus AI Assistant is a built-in coding and data agent designed for developers and database administrators using the TablePlus GUI. It assists users by writing, explaining, and executing SQL queries directly within the database workspace. The agent can autonomously query the database to gather context, though it requires explicit user approval for any non-read-only operations. It supports a wide range of backends, including cloud-based LLMs like OpenAI and Anthropic, local models via Ollama, and third-party agents through the Agent Client Protocol (ACP). By integrating directly into the right sidebar of the application, it streamlines the workflow of managing relational and NoSQL databases.
Query Assistance
- Query generation: The assistant can write complex SQL queries based on natural language prompts provided by the user.
- Query explanation: It can take existing SQL code and provide a detailed explanation of what the query does.
- Context awareness: Users can add specific tables and views as context to help the assistant understand the database schema.
- Direct execution: The assistant provides buttons to either copy generated code or run it directly in the SQL editor.
Database Interaction
- Autonomous querying: The assistant can query the current database by itself to check data before providing an answer.
- Read-only execution: Read-only queries like SELECT statements run directly without requiring manual intervention.
- Approval workflow: Any query that modifies the database requires the user to click an 'Approve SQL' button after reviewing the code.
- Safe mode integration: The assistant respects the application's safe mode settings even after a query is approved.
Provider Integration
- Cloud providers: Supports major LLM providers including OpenAI, Anthropic, Google AI, OpenRouter, and DeepSeek via API keys.
- Local models: Integrates with Ollama to run models locally on the user's machine without an API key.
- GitHub Copilot integration: Connects to GitHub Copilot for users with an active GitHub subscription.
- ACP support: Supports any coding agent that follows the Agent Client Protocol (ACP), such as Claude Code or Gemini CLI.
Specifications
- Deployment
- In-app sidebar (macOS)
- Supported Protocols
- Agent Client Protocol (ACP), Model Context Protocol (MCP)
- Local Execution
- Ollama, Codex CLI
- Cloud Integrations
- OpenAI, Anthropic, Google AI, OpenRouter, DeepSeek, GitHub Copilot
Added by AntSpace from public sources (tableplus.com).