Sim | Airweave

Sim is an open-source platform for building and deploying AI agent workflows visually. The native Airweave block lets you search your synced data collections from any workflow, no code required.

Prerequisites

Before you start you’ll need:

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.

Setup

Add the Airweave block

Open a workflow in Sim and drag the Airweave block from the tools palette onto the canvas.

Enter your Collection ID

Paste the readable ID of the Airweave collection you want to search. You can find this in the Airweave dashboard on the collection detail page.

Enter your API Key

Paste your Airweave API key into the API Key field. The value is stored securely and masked in the UI.

Write your Search Query

Type a natural-language query into the Search Query field. This field also accepts dynamic references from upstream blocks, so you can pass in user input or output from a previous step.

Configure search options (optional)

Adjust the retrieval strategy, max results, query expansion, reranking, and answer generation toggles to fine-tune your search.

Connect outputs to downstream blocks

Wire the block’s outputs, the results array and optional completion string, into downstream blocks such as an LLM agent, a Slack notification, or a condition filter.

Configuration

Parameter	Type	Default	Description
`collectionId`	string	-	The readable ID of the Airweave collection to search
`query`	string	-	Search query text
`apiKey`	string	-	Your Airweave API key
`limit`	number	25	Maximum number of results to return (10, 25, 50, or 100)
`retrievalStrategy`	string	`hybrid`	Retrieval strategy: `hybrid`, `neural`, or `keyword`
`expandQuery`	boolean	false	Generate query variations for better recall
`rerank`	boolean	false	Rerank results using an LLM for improved relevance
`generateAnswer`	boolean	false	Generate a natural-language answer from the search results

Output Format

The Airweave block returns two outputs:

results - an array of search result objects, each containing:

Field	Type	Description
`entity_id`	string	Unique identifier for the result entity
`source_name`	string	Name of the data source (e.g. “GitHub”, “Slack”)
`md_content`	string	Markdown-formatted content of the result
`score`	number	Relevance score
`metadata`	object	Additional metadata associated with the result
`breadcrumbs`	array	Navigation path to the result within its source
`url`	string	URL to the original content

completion - an optional string containing an AI-generated answer to the query. Only returned when Generate Answer is enabled.

Example Use Cases

RAG-style question answering: Connect the Airweave block to an LLM agent block. Pass the results or completion output as context so the agent can answer questions grounded in your synced data.
Automated notifications: Route search results to a Slack or email block to alert your team when specific content is found in your collections.
Relevance filtering: Chain the Airweave block with a condition block to filter results by score, then pass only high-relevance results to the next step.