For the complete documentation index, see llms.txt. This page is also available as Markdown.

Kanban Content Pipeline

The Kanban Content Pipeline is Sage's visual command center — every article in flight, every stage, every assignee, across every client.

The Kanban Board is Sage's operational view for content in flight. Where the Content Calendar shows you what's scheduled and the Performance Dashboard shows you what's published, the Kanban shows you what's moving through the pipeline right now — and who's responsible for each piece.

It's the page you use when you're actually running the day-to-day work.

Key Takeaways

  • Visual column-based view of every article's current stage

  • Drag-and-drop to move articles between stages (queued, generating, ready for review, published)

  • Filter to one client or view across all domains in your portfolio

  • Per-card actions: generate, preview, publish, edit — no need to navigate elsewhere

  • Replaces the messy "spreadsheet + email + Slack" workflow most agencies fall into

What the board shows

Each column represents a stage in the content pipeline. Each card is one article — drag them between columns or use the per-card actions to move them.

The columns

Column
Meaning
Card actions

Queued

Scheduled for generation but hasn't started yet

Edit title, generate now, delete

Generating

Sage AI Writer is running through the 7-stage pipeline

Watch progress, view stage timeline

Ready for Review

Generation complete, awaiting your approval before publish

Preview, edit, publish, regenerate

Published

Live on the client's WordPress site

View on site, view stats, edit

Failed

Generation hit a terminal error

Retry, view error, edit + retry

What each card shows at a glance

Element
What it tells you

Title

The article headline

Domain badge

Which client this article belongs to (domain name + favicon)

Status pill

Current stage in the pipeline

Source pill

Sage AI (auto-suggested) vs Manual (custom topic you added)

Owner initials

Who's assigned (writer, editor, reviewer)

Created/Updated time

How long ago this card last moved

Click any card to expand the full article view: preview, edit, full pipeline timeline, publish controls.

Filtering + organizing

The most common views:

  • One client at a time — pick a domain from the filter to see only that client's pipeline. Useful for daily standups or client check-ins.

  • All clients — see the full operational picture. Useful when checking overall agency throughput or hunting for stuck work.

  • Just the failures — filter to the Failed column to triage anything that went wrong. Sage's 30-minute stale-guard catches stuck generations, so the Failed column should be small and actionable.

Typical workflow

In review-first mode: you click Publish on each Ready-for-Review card to push to WordPress.

In autopilot mode: Sage skips the Ready-for-Review stop and pushes directly to Published — cards move Queued → Generating → Published without your involvement.

You can change modes per-domain at any time.

Per-card actions

Click any card to open the per-card actions:

Action
What it does
When to use

Preview

Opens the assembled article in a preview pane

Confirming an article reads right before publishing

Edit

Opens the article in the TinyMCE editor

Tweaking body, title, or meta before publish

Publish

Pushes the article to WordPress

When Ready for Review and approved

Generate now

Skips the schedule and runs the writer immediately

When you need an article faster than the scheduled date

Regenerate

Re-runs the writer on the same topic

When you don't like the current output

View stages

Shows the per-stage timeline (Brief → Keywords → Research → Writing → Image → Enriching → Publishing)

Diagnosing where generation got stuck

Delete

Removes the article from Sage

When the topic is no longer relevant

Why the Kanban exists

Most content agencies fall into one of two anti-patterns:

  1. The spreadsheet approach — a Google Sheet with article statuses, manually updated, perpetually out of date

  2. The chat-thread approach — Slack DMs and email threads with file attachments, no single source of truth

Both fail when you scale beyond 3-4 clients. The Kanban replaces both with a real-time view that anyone on the team can open and instantly understand: what's happening right now, what's stuck, what needs attention.

It's the difference between an operation that scales and one that doesn't.

When to use the Kanban

  • Daily morning check — see what moved overnight, what's ready for review, what failed

  • During team standups — visual overview of the week's pipeline by client

  • Hunting bottlenecks — articles stuck in Ready-for-Review for >7 days = reviewer is the bottleneck

  • Batch publishing — work through Ready-for-Review cards one at a time on a scheduled "publish day"

  • Client transparency — many agencies share read-only Kanban access with clients so they can see their pipeline

Frequently asked questions

Can clients see the Kanban?

Currently no — the Kanban is agency-side only. Client-visible status views are on the roadmap.

What happens if I drag a card to the wrong column?

For most moves: nothing destructive. You can drag back. But: dragging a Queued card to "Published" doesn't actually publish it to WordPress — that requires the explicit Publish action. The drag updates Sage's internal status only.

Can I assign articles to specific team members?

Currently the owner is derived from who queued the article or who generated it. Custom assignee/team-member features are on the roadmap. For now, use the article's title prefix or tags to indicate ownership manually.

What's the difference between Kanban and Content Calendar?

  • Content Calendar = scheduling view (when articles publish, calendar grid)

  • Kanban = operational view (current pipeline stage, who's doing what now)

Calendar answers "what's coming up?"; Kanban answers "what's happening right now?"

What's the "Failed" column for?

Articles that hit a terminal error during generation or publishing. Common causes: AI service hiccup, WordPress credentials revoked, source page timed out during research. Each failure surfaces a one-click retry; most resolve themselves on retry.

Last updated