Forage Glossary

4 min read · Last updated 10 November 2026 · Page version 10

Terms as Forage uses them.

The estate

Estate - the files Forage covers: SharePoint Online content and the content Squirrel has archived from it, as recorded by Squirrel.

Live - content still in SharePoint Online.

Archived - content Squirrel has moved into your own Azure storage.

Pointer page - the small page Squirrel leaves in SharePoint in place of an archived document. Anything looking at SharePoint sees the pointer page, not the document.

Forage for Copilot - the way Microsoft Copilot searches Forage and reads the text of the files Forage has read, archived files included. A Forage records administrator sees every file Forage holds; a Forage user sees only the files they can open in SharePoint.

Site Access - the record of who can open each SharePoint site Forage reads, as SharePoint's own permissions say, checked each night.

Shared items only - people who can open a file, folder or library inside a site because it was shared with them, and not the rest of the site.

What Forage has done with a file

Tracked - Forage knows the file exists, from Squirrel's record.

Indexed, or searchable - Forage has read the file's text and it can be found by search.

Excluded - Forage has not read the file, for a recorded reason: a type it does not read, or a type an administrator has switched off.

Lost - Forage tried to read the file and could not, with the cause recorded.

Read in full - every word in the file was read, so it can be found by what it says.

Partly read - most of the file was read, but part of it could not be, for example a damaged section or an attachment that would not open. The rest can be found by what it says.

Empty file - the file has no text at all, such as a blank document. Forage checked, and nothing is missing.

No text found - Forage read the file but found no words it could recognise. These are mostly photos, drawings and scans. Unlike an empty file, a picture can hold words Forage could not make out.

Unread scans - files whose scanned pages were not read. Their typed text can be found in search, but the words in their scans cannot yet. They are read again when the file next changes.

Coverage - how much has been read, reported against the files in the sites you include, with every SharePoint site shown beside it for reference, rather than against what was convenient to read.

Checked - the file has been checked against the content rules active now. Checked is not the same as indexed: a document can be searchable before it has been checked, and when the rules change the checked count falls and refills.

Indexed and searchable never mean classified, and classified never means that a retention decision has been made.

Content classification

Content label - a kind of content a rule recognises, such as payment card numbers or contract terms.

Severity - how serious a label is treated as being.

Content rule - the rule that produces a label, built from phrases, patterns, checksum validation and proximity.

Finding - one document matching one rule, keeping the words that produced it.

Suggested label - a label the AI assistant proposes from groups of similar files it has read. Nothing is labelled until a person accepts one and switches its rule on.

Test run - testing a rule before it goes live. It shows the most the rule could match, a sample checked with the real rule, and a projected range across the documents read so far.

Version, draft, active, retired - every change to a rule is a new version rather than an edit in place. A draft that has never been used can be deleted; an active rule can be retired.