- LISTS
- VALIDATION
- REPORTS
- HUMAN
Short answer
How can I audit document metadata using my AI?
Your AI uses watsonx Discovery to list all available collections and pull specific metadata fields from individual documents. This lets you spot missing tags or incorrect properties without manual searching. You'll see exactly how the AI navigates your data structures below.
Audit outcomes
Where your data issues land.
What you get after the AI finishes its scan.
LISTS
Collection inventory
The AI identifies every collection in your instance. You get a clear view of your data organization.
VALIDATION
Metadata integrity
The AI checks specific document properties against your requirements. It flags files that don't match your schema.
REPORTS
Error summaries
Your AI compiles a list of documents with broken or missing metadata. This makes cleanup straightforward.
HUMAN
Manual fixes
The AI identifies the issues, but you decide how to correct the source data. It won't change your files for you.
The workflow
What your AI does when the request arrives.
The AI moves through your Discovery instance step by step to find what you need.
Map the environment
The AI first finds all the data groups available to it in your Discovery instance.
list_discovery_collectionsScan document lists
Once it knows which collection to look at, the AI pulls a list of all documents within that group.
list_collection_documentsInspect specific files
The AI drills down into a single document to read its full metadata profile.
get_document_detailsAnalyze and report
The AI compares the retrieved metadata against your audit rules to find discrepancies.
get_document_details
Try it
Copy these to start.
Use these prompts to kick off an audit.
Starting points
Swap out the collection names and document IDs for your actual Discovery data.
watsonx Discovery Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_MOvCoAeU4n0XLtoAj5AMd7HKRadaRJgAFtxeD6Bh/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — watsonx Discovery capabilities are ready to use.
{
"mcpServers": {
"watsonx-discovery-mcp": {
"url": "https://edge.vinkius.com/vk_preview_MOvCoAeU4n0XLtoAj5AMd7HKRadaRJgAFtxeD6Bh/mcp"
}
}
}List all my Discovery collections so I can see what's available.
Show me all the documents in the 'legal_contracts' collection.
Get the metadata details for document ID doc_98765 to check its timestamp.
Check the metadata for all documents in the 'user_logs' collection and tell me if any are missing a 'source' tag.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect watsonx Discovery once, then ask.
Follow the one-click setup to link your instance. Your credentials stay encrypted, and you can immediately start asking questions about your data.
Connect watsonx Discovery to your AIFAQ
How this task behaves
- 01
Can the AI change my document metadata?
No. The AI can only read and inspect metadata using the available tools. You must update the metadata manually in watsonx Discovery.
- 02
Can the AI delete documents from my collections?
No. There are no tools provided to delete or modify data, so your files remain untouched.
- 03
Does the AI see all my data?
The AI only sees the data within the collections it can access through the provided tools.
- 04
Can I use this to find missing tags?
Yes. By pulling document details, the AI can identify which files are missing specific metadata fields.
- 05
How does the AI know which collection to audit?
You tell the AI which collection to look at, or you can ask it to list all collections first to find the right one.
More questions about watsonx Discovery? The Connector page answers them. See everything the watsonx Discovery Connector can do
Connect watsonx Discovery to Claude, Cursor, ChatGPT & more
