How to Connect Google Antigravity to fanTask via MCP: Step-by-Step Guide

What is the Model Context Protocol (MCP) and how does Antigravity use it?

The Model Context Protocol (MCP) is an open standard designed to enable Large Language Models and autonomous coding agents to communicate securely and directly with external services, developer tooling, and structured data sources. Instead of isolating an AI agent within the boundary of local directory files or relying on manual prompt copy-pasting, MCP provides a standard JSON-RPC interface that agents can query and invoke dynamically.

In Google Antigravity — whether running headless or interactive workflows through the agy command-line interface or developing inside the Antigravity IDE —, MCP integrations bridge the gap between high-level model reasoning and real-world project infrastructure.

Why connect Google Antigravity to fanTask?

Google Antigravity provides powerful agentic capabilities, including multi-agent orchestration, iterative self-correction, and tool execution. However, like all advanced AI coding assistants, it begins each new conversation or session cold. Without an external, durable memory layer, agents risk re-introducing regressions, straying from architectural conventions, or repeating work that was already completed or rejected.

fanTask provides an MCP-native project and task management backbone specifically designed for human-AI software engineering teams. Connecting Google Antigravity to fanTask gives your agent critical superpowers:

• It reads your living architectural guidelines, gotchas, and constraints (via the persistent AI Context field) before touching code.
• It searches past task discussions and resolution records to inspect how previous bugs and edge cases were solved.
• It synchronizes work transparently on a shared board, keeping statuses updated and attaching structured audit comments with affected files and test outcomes.

Step 1: Get your access token from fanTask

If you just created your account, fanTask automatically shows you a "Connect your AI to fanTask" modal right on your workspace page, with a few options (Claude, Gemini CLI / Antigravity, ChatGPT, Gemini App, Other). Pick Gemini CLI / Antigravity — the platform mints a personal access token on the spot and hands you a ready-to-paste setup prompt for an Antigravity or agy session.

If you closed that modal, or you're on an existing account, you can pick this back up anytime from your workspace menu: Workspace → MCP / AI. It offers the same options and the same ready-to-paste prompt, generated with your current token.

Step 2: Register the fanTask server in Antigravity via the agy CLI

Google Antigravity provides a direct, agent-actionable command-line interface via the agy CLI tool. Rather than clicking through manual GUI preferences, you can register the fanTask MCP server using the following documented command:

agy mcp add --header "Authorization: Bearer YOUR_FANTASK_API_TOKEN" fantask https://api.fantask.ro/mcp

Replace YOUR_FANTASK_API_TOKEN with the personal access token you generated in Step 1. This command registers an MCP server alias named fantask, configures the Bearer token authorization header, and points traffic to the production endpoint at https://api.fantask.ro/mcp over encrypted TLS.

Step 3: Restart session and verify available tools

A crucial architectural rule of Model Context Protocol clients is that MCP servers are loaded strictly at client startup. After executing the registration command, you must restart your Antigravity IDE or start a fresh session in the agy CLI to initialize the connection.

Once the new session initializes, Antigravity discovers the fanTask MCP server and gains access to the full suite of project coordination tools:

• fantask.get_project_context(project_id): Loads your project's living architectural boundary document, including forbidden libraries, environment quirks, and testing requirements.
• fantask.search_tasks(query, filters): Executes lexical and semantic searches across task histories, comments, and past resolutions to leverage prior knowledge.
• fantask.get_task(task_id): Fetches complete specifications, requirements, and comments for an assigned ticket.
• fantask.update_task(task_id, status, comment): Transitions task statuses (e.g. moving an item to in_progress or in_review) and documents progress.
• fantask.create_task(project_id, title, description, priority): Automatically creates follow-up tickets or subtasks when unexpected technical debt or bugs are discovered during implementation.
• fantask.add_comment(task_id, message): Appends detailed technical audit notes, specifying files changed, line ranges touched, and rationale for review.

Recommended Collaborative Workflow

Integrating Google Antigravity with fanTask establishes a clear, repeatable engineering loop:

1. Session Initialization: Begin your session by instructing the agent: "Review our project AI Context in fanTask and list our active tasks." Antigravity calls get_project_context and search_tasks, immediately anchoring its reasoning in your established engineering constraints.
2. Implementation: As Antigravity starts a task, it calls update_task to set the status to in_progress. If it encounters ambiguous behavior or tricky dependency conflicts, it uses search_tasks to see if a workaround was already documented.
3. Review & Hand-off: When implementation and verification tests are complete, Antigravity moves the ticket to in_review and posts an audit summary via add_comment. You inspect the git diff, verify the changes, and mark the task as done yourself.

Get Started for Free

The fanTask Free plan provides 1 active project board, up to 3 team members, persistent AI Context, and full MCP tool access. Connect Google Antigravity today and give your AI agent durable project memory.

Create a free fanTask account and connect Google Antigravity →