• VALIDATION
  • METADATA
  • MISSING
  • HUMAN

How to audit document indexing with your AI

Connect Haystack to your AI client to inspect how your files are stored and organized. Once connected, your agent can pull specific file details to confirm your data is ready for use.

Ask AI about this page

Short answer

How do I check if my files are indexed correctly?

Your AI can list your Haystack workspaces and files to verify their presence. It can then pull specific file metadata to confirm the indexing matches your expectations. You'll see exactly what the system sees.

Where every issue lands.

Audit outcomes

Your AI categorizes the results of your inspection into these areas.

VALIDATION

Confirmed file presence

The AI uses file lists to prove your documents actually exist in the Haystack environment. You get a clear list of what's live.

METADATA

Verified file properties

Your agent inspects the specific attributes of a file to ensure the indexing captured the right information. This confirms the data is usable.

MISSING

Identified indexing gaps

If a file is missing from the workspace list, your AI flags it immediately. This helps you find where the ingestion process failed.

HUMAN

Manual fixes needed

When the AI finds a file with incorrect metadata, it flags the discrepancy for you. You'll need to manually adjust the source or re-run the indexer.

The workflow

What your AI does when the audit starts.

The AI moves through your Haystack structure to find and inspect your data.

  1. Locate workspaces

    The AI identifies all available Haystack workspaces to establish the scope of the audit.

    list_workspaces

  2. Scan for files

    It pulls a list of every file within a specific workspace to check for missing documents.

    list_files

  3. Inspect file details

    The agent pulls the specific metadata for a single file to verify its properties.

    get_file

  4. Report discrepancies

    The AI compares the actual file data against your requirements and highlights any errors.

    get_file

Try it

Copy these to start.

Use these prompts to begin your audit.

Starting points

These are starting points. Swap out the workspace or file names for your actual Haystack identifiers.

deepset Cloud Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_Tf2g8zLZroyDlliFU39whBZWkEDykSKQVz38t3zr/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — deepset Cloud capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "haystack-deepset-cloud-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_Tf2g8zLZroyDlliFU39whBZWkEDykSKQVz38t3zr/mcp"
    }
  }
}
Copy into chat04

  • List all my Haystack workspaces so I can see where my files are stored.

  • Show me all the files in the 'production_docs' workspace.

  • Get the metadata for the file named 'user_manual_v2.pdf' to check its index status.

  • Check the 'marketing_assets' workspace and tell me if any files are missing from my expected list.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Haystack once, then ask.

Just link your Haystack account once. Your credentials stay encrypted, and you can immediately start asking your AI about your files.

Connect Haystack to your AI

FAQ

How this task behaves

  • 01

    Can the AI delete files in my Haystack workspace?

    No. The AI can only list and get file information. It cannot delete or modify your files.

  • 02

    Can the AI fix broken indexing automatically?

    No. The AI can identify where the indexing is wrong by checking metadata, but you must perform the actual fix in Haystack.

  • 03

    How does the AI know which files are in which workspace?

    The AI uses the list tools to map out the workspace hierarchy before inspecting individual files.

  • 04

    Can I see the actual content of the files?

    Yes, the AI can access file details through the available tools to help you verify the content is indexed.

  • 05

    Does the AI have access to all my Haystack data?

    The AI only has access to the data it can reach through the specific tools provided in this MCP.

  • More questions about Haystack? See everything the Haystack Connector can do

Connect deepset Cloud to Claude, Cursor, ChatGPT & more