The Library: every article across your projects, in one table

The Library is the audit view of your content. Where the Desk shows only what needs a decision, the Library shows every article in every project, at every stage, so you can see where each one stands and which have stopped moving. It is a read and filter screen: there are no bulk actions on it.

What the Library shows#

The header counts the articles and the projects they belong to. The table has one row per article, with six columns:

• Title, linking to the article's detail page in its content plan.
• Project.
• Status: Planned, Generating Brief, Brief Ready, Drafting, Draft Ready, Reviewing, Ready to Publish or Published.
• Source: radar, research, idea or manual, by where the article came from.
• Distribution: one chip per channel the article went out on, such as LinkedIn ×2 · Jul 14, with the count and the latest date. A chip dated in the future is scheduled outside VarynForge and not live yet. An article with a planned publish date and nothing attested shows a scheduled Article chip. Five channels covers the channels.
• Age in state: how long since the article last changed status.

A row is marked stuck when it has sat in one status longer than that status should take. Hovering the mark says how many days. The limits are 1 day in Generating Brief, 3 days in Draft Ready or Reviewing, 7 days in Brief Ready or Ready to Publish, and 14 days in Drafting. Planned and Published articles are never stuck.

Where the data comes from#

The Library covers the projects of every organization you belong to. It reads the article suggestions in your content plans. Age in state comes from the article's history: the time of its latest status change, or its last update when no change is recorded. Distribution comes from the distribution ledger: a row when an article is attested as published, and a row for each derived asset, such as a social post, registered against it with register_derived_asset. Everything is read when you open the screen.

What you can do here#

• Status, Source and Project dropdowns filter the table. Status and Source take several values at once; Project takes one, or All projects.
• Stuck only keeps only the rows marked stuck.
• Clear filters resets all four.
• The Title, Project, Status and Age in state headers sort the table; a second click reverses the order. The default puts the oldest change first.
• Load more fetches the next 200 rows when there are more.

Filters are kept in the page address, so a filtered view can be bookmarked. Sorting and Stuck only act on the rows already loaded. With no articles yet, the Library links back to the Desk.

The same rows over MCP#

list_article_suggestions returns one project's articles with the same fields: title, status, source, publishedAt, scheduledFor, and a distributions rollup of { channel, count, latestAt, scheduledFor } per channel. Your agent pages through it per project to build the cross-project table, and reads a missing channel in distributions as a gap, for example an article with no linkedin entry has no LinkedIn post yet. get_article_suggestion returns one article's full record, and update_article_status moves it on. Age in state and the stuck flag are computed on this screen only; no tool returns them.