Archive and drop the retired page content columns

[RaviTejaKNTS]

Why

This is release 2 of plan step 2e. Release 1 (#42, live in 72f5bec7) removed every reader of the old page content columns and made them optional. Nothing on the site reads them now, so they can go.

What changed

• Migration 20261010063116_retire_page_content_columns.sql:
  • copies every value of the retired tool, directory and catalog content columns into public.stack127_retired_page_content, which has no public read access;
  • then drops the columns.
  • It's idempotent. It was applied in development, where the archive held 902 values, no live columns were left and no public function read them. Running it again keeps the archive and the page dates intact.
• Plan doc: steps 2e and 3 are marked done.

Checks

• Sol reviewed the migration in two rounds, as part of Finish plan step 2e and clear the content pass follow-ups #42's review.
• The release 1 compatibility test passes with the columns present and absent.
• Production is on 72f5bec7, and its readers don't use the columns.

🤖 Generated with Claude Code