### Markdown Structure for Agent Skill Instructions Source: https://github.com/alphaperseii3000/google-antigravity-docs/blob/master/google-antigravity-docs.md Provides the main content structure for a SKILL.md file, including detailed instructions, usage scenarios, and step-by-step guidance for the agent. ```markdown # My Skill Detailed instructions for the agent go here. ## When to use this skill - Use this when... - This is helpful for... ## How to use it Step-by-step guidance, conventions, and patterns the agent should follow. ``` -------------------------------- ### YAML Frontmatter for Agent Skill Source: https://github.com/alphaperseii3000/google-antigravity-docs/blob/master/google-antigravity-docs.md Defines the metadata for an agent skill, including its name and a brief description. This is a required component at the top of the SKILL.md file. ```yaml --- name: my-skill description: Helps with a specific task. Use when you need to do X or Y. --- ``` -------------------------------- ### Agent System Configuration and Modes Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet outlines the configuration options and operational modes for the Agent System in Google Antigravity. It covers conversation modes (Planning and Fast), the location of global and workspace rules, and terminal command auto-execution policies. ```bash # Agent conversation modes # Planning Mode: For deep research, complex tasks, or collaborative work # Fast Mode: For simple tasks that can be completed quickly # Agent settings location # Global rules: ~/.gemini/GEMINI.md # Workspace rules: .agent/rules/ # Terminal command auto-execution policies # - Request Review: Never auto-execute (except allowlist) # - Always Proceed: Never ask for review (except denylist) ``` -------------------------------- ### Antigravity Editor Keyboard Shortcuts and Features Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet details essential keyboard shortcuts for navigating and interacting with the Antigravity Editor, which is based on VS Code. It also highlights key 'Tab' features like Supercomplete, Tab-to-Jump, and Tab-to-Import for enhanced coding efficiency. ```bash # Keyboard shortcuts # Open Agent Manager: Cmd + E (Mac) / Ctrl + E (Windows) # Open Settings: Cmd + , (Mac) / Ctrl + , (Windows) # Trigger Command: Cmd + I (Mac) / Ctrl + I (Windows) # Toggle Terminal: Cmd + J (Mac) / Ctrl + J (Windows) # Open Quick Picker: Cmd + P (Mac) / Ctrl + P (Windows) # Tab features # - Supercomplete: File-wide code suggestions # - Tab-to-Jump: Navigate to next logical edit location # - Tab-to-Import: Auto-add missing import statements ``` -------------------------------- ### Available Reasoning Models in Antigravity Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet lists the frontier models available in Google Antigravity for core reasoning tasks, sourced from the Google Vertex Model Garden. It also mentions additional specialized models used for specific functionalities like image generation and browser automation. ```bash # Available reasoning models: # - Gemini 3 Pro (high) # - Gemini 3 Pro (low) # - Gemini 3 Flash # - Claude Sonnet 4.5 # - Claude Sonnet 4.5 (thinking) # - Claude Opus 4.5 (thinking) # - GPT-OSS # Additional specialized models (not user-configurable): # - Nano Banana Pro: Generative image tool # - Gemini 2.5 Pro UI Checkpoint: Browser subagent actuation # - Gemini 2.5 Flash: Checkpointing and context summarization # - Gemini 2.5 Flash Lite: Codebase semantic search ``` -------------------------------- ### MCP Server Configuration in Antigravity Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet explains how to configure custom MCP (Model Context Protocol) servers in Google Antigravity for secure integration with various external services. It specifies the file location for the configuration and lists supported integration categories. ```json // Custom MCP server configuration location: // Access via "..." dropdown > MCP Store > Manage MCP Servers > View raw config // File: mcp_config.json // Supported integrations include: // - Databases: BigQuery, PostgreSQL, MySQL, MongoDB, Redis, Spanner // - DevOps: GitHub, Heroku, Netlify, Firebase // - Project Management: Linear, Notion, Atlassian, Dart // - Design: Figma Dev Mode // - Payments: Stripe, PayPal // - AI/ML: Perplexity Ask, Pinecone ``` -------------------------------- ### Antigravity Skills Framework Structure and Location Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet details the structure and file locations for the Skills Framework in Google Antigravity. Skills are reusable packages that extend agent capabilities, and this documentation specifies where to place them globally or per workspace, and the required format for the SKILL.md file. ```markdown --- name: my-skill description: Helps with a specific task. Use when you need to do X or Y. --- # My Skill Detailed instructions for the agent. ## When to use this skill - Use this when... ## How to use it Step-by-step guidance for the agent. ``` -------------------------------- ### Antigravity Browser Integration Capabilities and Security Source: https://context7.com/alphaperseii3000/google-antigravity-docs/llms.txt This snippet describes the capabilities of Antigravity's browser integration, including controlling Chrome for testing and data retrieval. It also outlines the security measures, such as denylist/allowlist configurations and the use of a separate Chrome profile for isolation. ```bash # Browser subagent capabilities: # - Clicking, scrolling, typing # - Reading console logs # - DOM capture, screenshots, markdown parsing # - Video recording of actions # Browser security settings: # - Denylist: Server-side check via Google Superroots BadUrlsChecker # - Allowlist: Local text file for trusted URLs # - Runs in separate Chrome profile for isolation # Chrome profile location setting: # Configurable in Settings > Browser section ``` === COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.