# Unpublished Content Report

> Create a live HTML governance report of unpublished or draft documents and site pages that haven't been edited for a specified number of days, sorted with the stalest items first, and save it in a SharePoint report library. Use when the user says: - "Create a report of unpublished documents and pages" - "Show stale draft content on this site" - "Find content not published and not edited for 14 days" - "Generate an unpublished content report" - "Save a report of old drafts in the Reports library"

- Skill: `pnp/unpublished-content-report` (Agent Skill)
- Install (CLI): `npx skillmds@latest add pnp/unpublished-content-report`
- Raw SKILL.md: https://api.skillmd.com/api/skills/pnp/unpublished-content-report/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Web & Frontend
- Author: pnp (https://skillmd.com/u/pnp)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/pnp/unpublished-content-report

---

# Unpublished content report

## When to use
Use this skill when the user wants an HTML governance report covering SharePoint documents and site pages that aren't published and haven't been edited for a specified period. Use it for stale drafts, pending/rejected content, checked-out pages, or unpublished changes. The report must sort by days since last edit in descending order and be saved in a document library.

## Inputs
- Current SharePoint site URL and context.
- Inactivity threshold in days. Default to 14 only when the user doesn't provide one.
- Destination document library. Default to `Reports` only when the user doesn't provide one.
- Optional report filename. Otherwise use `Unpublished-content-inactive-<days>-days.html`.
- Scope is documents in user-facing document libraries plus pages in the site's Site Pages library. Exclude system, asset, style, preservation, template, report, and `AgentAssets` libraries unless the user includes them.

## Steps
1. Get the current date/time before calculating relative ages.
2. Discover the site's lists and libraries, including the Site Pages library. Identify user-facing document libraries and exclude system libraries.
3. Verify the destination library exists. If it doesn't, create it as a document library with a clear report-oriented description.
4. Retrieve schemas before querying items. Identify these internal fields where available: `FileLeafRef`, `FileRef`, `Modified`, `Editor`, `_ModerationStatus`, `_UIVersionString`, `CheckoutUser`, and page fields such as `Title`, `FirstPublishedDate`, and `PromotedState`.
5. Query all files recursively with a row limit equal to the known item count, or 50,000 when unknown. Retrieve the complete dataset before filtering.
6. Determine publication state from SharePoint metadata rather than filenames:
   - In moderated libraries, treat `_ModerationStatus` values equivalent to Approved/Published as published. Pending, Draft, Rejected, or other non-approved values are unpublished.
   - In versioned libraries, a version ending in `.0` is a published major version. A version ending in a non-zero minor number indicates draft or unpublished changes.
   - Treat checkout-only pages, never-published pages, and pages with draft changes as unpublished.
   - Don't classify an item as unpublished solely because the library doesn't use moderation or minor versions. If no reliable publication-state signal exists, omit the item and note the skipped library in the report.
7. Calculate whole days since `Modified`. Keep items whose age is greater than or equal to the requested threshold.
8. Normalize each match with item name and link, type, library, publication state, version, last edited date/time, editor, and days since last edit.
9. Sort by days since last edit descending, then item name ascending as a stable tie-breaker.
10. Build the HTML using `Report.html` and, when practical, `LiveData.html` so results refresh from source libraries when opened. Pass every source list identity to the live-data tool; don't hand-write REST or Graph calls.
11. Include summary cards for total items, documents, and pages; a concise criteria note; and a table containing all matching items. Escape user-controlled values before inserting them into raw HTML.
12. Route generated HTML through an output data reference, then create the `.html` file in the destination library using that reference. Don't reconstruct content from a truncated preview.
13. Verify the file was created and return a link. If any tool fails or data is incomplete, say so plainly and don't invent results.

## Output format
Return a concise completion message containing:
- A markdown link to the created HTML report.
- The threshold and destination library.
- A note that the report refreshes when opened if LiveData was used; otherwise state that it's a point-in-time report.
- Any libraries skipped because publication state couldn't be determined reliably.

The HTML report must contain:
1. Title: `Unpublished content inactive for <days>+ days`
2. Site and criteria statement
3. KPI cards: Total items, Documents, Pages
4. Table columns: Item, Type, Library, Publication state, Version, Last edited, Edited by, Days since edit
5. Rows sorted by Days since edit descending

Example request: `Create a HTML report showing documents and pages that aren't published and haven't been edited for 30 days. Save it in Governance Reports.`