> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.eyelevel.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.eyelevel.ai/_mcp/server.

# Use GroundX With Your Agent

Ask your connected agent to search documents, extract data, or build reports.
The examples below show three common tasks.

Before you start, install the harness and connect your agent to GroundX. See
[Install GroundX Agent Harness](/documentation/agent-harness/install-ground-x-agent-harness) and
[Connect Hosted MCP Tools](/documentation/agent-harness/connect-hosted-mcp-tools). Use a GroundX API
key that can add and search documents; a read-only key won't be able to load or
search anything. An agent that can't reach GroundX can explain the steps but
can't run them.

## Ask questions about your documents

Hand your agent a document and ask about it, all in plain language:

```text
Set up a GroundX bucket called "my-project", add https://example.com/report.pdf
to it, and tell me when it's ready to search.
```

```text
What were the Q3 revenue figures in that report, and which page do they come from?
```

The agent creates a bucket (a document container), loads the document, waits for
processing, then searches it to answer your question. Ask it to "show the exact passages" or
"cite the sources" for the supporting text and pages. If your agent can read
local files and upload them over HTTP, give it a file path instead of a web link.

## Pull structured data out of your documents

When you want a document to come back as structured data your application can
use, tell the agent the fields you want (or paste a sample of the JSON you want
back) and hand it a few example documents:

```text
Here are three utility statements: <links>. Pull the account number, due date,
total amount due, and service address from each one, as JSON.
```

The agent designs the extraction, runs it, and compares the results with the
documents. Review its output and explain any errors (for example, "use the
statement date instead of the payment date"). Once the results are accurate on
your samples, run it on the rest of your documents.

## Generate a report across a set of documents

When you have a whole set of documents and the same questions to answer about
each one, ask the agent to work through them and assemble a report:

```text
Go through every statement in the "claims" bucket, answer these questions for
each one: [your questions], and put the results in a report with the source for
each answer.
```

The agent answers your questions for each document and assembles a report with
links to the supporting passages. Use this for tasks such as reviewing
statements for fraud or summarizing contracts.

## Prefer to write this in your own code?

To build these tasks into your application, use the following guides:

* [Connect Hosted MCP Tools](/documentation/agent-harness/connect-hosted-mcp-tools) for the tools
  your agent can call.
* The [Ingest Tutorial](/documentation/guides/ingest-tutorial) and [Search Tutorial](/documentation/guides/search-tutorial) for loading and searching
  documents in code.
* [Extract Data from Documents](/documentation/guides/extract-data-from-documents) for the
  structured-data workflow at the SDK level.
* The [Direct SDK/API Quickstart](/documentation/get-started/direct-sdk-api-quickstart) to get
  started.

## Next Steps

* Something not behaving as expected? See [Verify and Troubleshoot](/documentation/agent-harness/verify-and-troubleshoot).