Antigravity Workspace FAQs

Question 1

How does Antigravity generate its codebase knowledge base?

Accepted Answer

It utilizes a dynamic multi-agent cluster that employs smart functional grouping to autonomously read and analyze your code. Each module gets its own agent that generates a detailed Markdown knowledge document, and a Map Agent creates a routing index for intelligent, precise question answering.

Question 2

Which IDEs and LLMs are compatible with Antigravity?

Accepted Answer

Antigravity Workspace offers broad compatibility, supporting popular IDEs like Cursor, Claude Code, Windsurf, and VS Code. It integrates with various LLMs, including Claude Code, Codex CLI, and Gemini CLI, via its MCP server for seamless interaction.

Question 3

What is Antigravity Workspace?

Accepted Answer

Antigravity Workspace is a multi-agent knowledge engine and MCP server that transforms any codebase into a queryable AI assistant. It provides grounded answers to code-related questions with precise references, effectively acting as a 'ChatGPT for your repository'.

Question 4

How does Antigravity Workspace improve AI agent capabilities?

Accepted Answer

It significantly enhances AI agent capabilities by autonomously building a comprehensive, high-quality knowledge base from your repository. This rich, contextual information reduces hallucinations and allows agents to provide more accurate, grounded responses based on your actual codebase.

Question 5

What are the core commands for using Antigravity?

Accepted Answer

The primary commands are `ag-refresh`, which builds the comprehensive multi-agent knowledge base from your codebase, and `ag-ask`, which allows you to query the codebase and receive grounded answers. An `ag-mcp` server is also available for LLM integration.

Antigravity Workspace

Antigravity Workspace

주요 기능

사용 사례

주요 기능

사용 사례