Admin / Editor · Allow about 7 minutes
Before you begin
Owner, Admin or Editor with Knowledge Intelligence access. Analysis availability depends on workspace capability, configured services and AI usage allowances.
1. Scope the inventory
Open the Knowledge Map or Knowledge Intelligence area and its Knowledge inventory view.
Filter by the relevant knowledgebase, category, status, content type, difficulty or analysis state. Apply the filters and inspect the item count.
Keep the scope aligned with your question: a filtered set is not a complete workspace coverage assessment.
2. Inspect evidence and freshness
Open an item to compare its article, classification and latest analysis state. Check whether the source changed after analysis.
Treat missing or failed analysis differently from a confirmed content gap. Read warnings and source evidence before making a recommendation.
Compare findings with the current article and intended reader task. Automated classification is a review aid, not approval to rewrite or publish.
3. Refresh deliberately
Select the intended eligible items for analysis and review the scope and any usage or capability message before starting.
Start the supported analysis action only when the scope and potential AI usage are acceptable. Track queued, running, complete and failed states rather than repeatedly requesting the same work.
Return after completion, review new findings and check their source freshness. Use the recommendation workflow for decisions; inventory refresh does not publish articles.
Check your result
The reviewed inventory has a clear scope and freshness, and any analysis request has a known state and usage impact.
If something goes wrong
Analysis unavailable: check the workspace capability, configured AI service and usage allowance with an Admin.
Stale finding: compare the current source, refresh the analysis where available and reassess the evidence.
Job failed or stalled: record the item, time and visible error; check service availability before starting repeated work.