Suggest Metadata Columns
Most SharePoint libraries start with little to no metadata or just the basics. Over time, this leads to disorganized content, inconsistent tagging, and difficulty finding the right documents. Our Suggest Metadata Columns skill analyzes the actual content of your documents to recommend metadata that reflects how your business really works. Instead of guessing what fields to create, you get a data-driven starting point based on real, recurring information across your files.
What you’ll get
- A recommended list of metadata columns based on actual document content
- Clean, consistent column names using concise PascalCase
- A suggested data type for each column (Text, Date, Choice, etc.)
- A short description explaining what each column captures
- An autofill prompt to support automated metadata tagging
- Grouped recommendations (e.g., Identifiers, Dates, Financial, Parties, Terms)
How it works
- Identify the current location
The skill determines which folder or document library you’re working in. - Check existing metadata
Current columns are reviewed to ensure no duplicate suggestions are created. - Scan documents
The skill evaluates the files in your library, including names, types, and metadata. - Analyze content deeply
It extracts all meaningful data points from documents—not just obvious fields like title or date. - Compile a master dataset
Data points are consolidated across documents, prioritizing recurring and high-value information. - Cross-check for completeness
A final validation ensures no important data points are missed. - Generate metadata suggestions
Each data point is turned into a proposed column with a name, type, description, and autofill logic. - Present recommendations
Results are grouped into logical categories and shared for review, refinement, and approval.
When to use this
- When your SharePoint library lacks meaningful metadata
- When you want metadata based on real document content (not assumptions)
- When existing columns are inconsistent or incomplete
- When planning a metadata structure before creating new fields
Output
A structured list of recommended metadata columns, grouped by category.
Each suggestion includes:
- Column name (PascalCase)
- Data type
- Description
- Autofill prompt
If a column requires predefined options (e.g., Choice), the output flags that confirmation is required before creation.
Why this matters
Metadata is what makes SharePoint useful but defining it manually is often incomplete or inconsistent. This skill takes a fundamentally different approach:
- It analyzes what your documents actually contain, not what you think they contain
- It captures all meaningful data points, not just the obvious ones
- It ensures metadata reflects real business patterns and usage
The result is better filtering, search, automation, and overall document organization—without starting from scratch.
