Google Antigravity

Google Antigravity is an agentic IDE designed specifically for agent-first coding workflows. It allows developers to rapidly create, test, and deploy multiple AI agents that collaborate with code, apps, and data. Airweave MCP is part of the first set of integrations available in the Google MCP Store, enabling your coding agents to instantly search across all your synced apps and databases with zero additional setup required.

Prerequisites

Before you start, you will need:

A Google Antigravity account: Sign up at antigravity.google
A collection with data: At least one source connection must have completed its initial sync. See the Quickstart if you need to set this up.
An API key: Create one in the Airweave dashboard under API Keys.

Adding Airweave from the MCP Store

Open Google Antigravity
Open the MCP Store
Search for Airweave
Click Add to enable the integration
Enter your Airweave API key when prompted

Once added, your agents can search across all your Airweave collections as part of their autonomous workflows.

Available Tools

After enabling the integration, your agents have access to the following tools:

`search-{collection}`

Search within your configured Airweave collection. The tool name includes your collection ID, for example search-my-knowledge-base.

Parameter	Type	Description
`query`	string	Your search query (required)
`limit`	number	Maximum results to return (default: 100)
`offset`	number	Number of results to skip for pagination (default: 0)
`response_type`	string	`"raw"` for results, `"completion"` for AI-generated answers
`recency_bias`	number	Weight recent content from 0.0 to 1.0 (0 = no bias, 1 = recency only)
`score_threshold`	number	Minimum similarity score from 0.0 to 1.0
`search_method`	string	`"hybrid"`, `"neural"`, or `"keyword"`
`expansion_strategy`	string	`"auto"`, `"llm"`, or `"no_expansion"`
`enable_reranking`	boolean	Use LLM-based reranking (default: true)
`enable_query_interpretation`	boolean	Extract filters from natural language (default: true)

`get-config`

Display the current Airweave MCP server configuration, including the configured collection and available commands.

Example Usage

Once Airweave is added, your agents can search your data using natural language:

Search my Slack workspace for discussions about the Q4 roadmap

Find all documents mentioning the pricing update from last week

What were the key decisions from yesterday's standup? Search with recency bias.

Search Capabilities

The Airweave MCP integration provides:

Hybrid search: Combines semantic and keyword matching for best results
Query expansion: Generates variations of your query for better recall
Reranking: Applies LLM-based reranking for improved relevance
Context retrieval: Returns relevant context with breadcrumbs to original sources
AI-generated answers: Optionally returns synthesized answers from search results

Self-Hosted Instances

If you are running a self-hosted Airweave instance, you can configure a custom base URL in the MCP settings after adding Airweave from the store.