The hard part of engineering is often not the new build. It is keeping what already exists working safely. In a content system, old pages are that existing plant.

Aging and expiry happen together. Names, prices, interfaces, and rules change while the page is still cited. The risk is not "no new article." It is an old article restated as a stale fact. Give each critical fact an expiry: a date, a version, or the update of a source it depends on. If it is not rechecked, remove it from the citable stock or mark it as history.

Capacity is the scale you can actually maintain. The stricter the standard, the fewer pages you can keep honest per week. A standard beyond capacity gets bypassed under deadline, which means there is no standard. Limit the number of citable pages by how many facts you can recheck and how many rejections you can clear in a week.

Write the failure plan before the failure. If a source is shown to be unreliable, which pages must come down the same day. If an engine starts restating a wrong boundary, do you change the fact base first or the highest-traffic page. Without a plan, everyone edits a different sentence and adds contradictions.

The future of the content system is not unlimited addition. It is keeping the citable stock true within capacity and risk. A new page qualifies when it does not set the old stock against itself.