v0.13.41: @newsroom.writer: Capture first, and the PHP story's own site on the front page

Capture first: the newsroom's first story covered while it was breaking. A BBC live page at breakfast became an evidence vault at 08:03 UTC, a section here at 08:13, and a site of its own at 09:27. By 12:05 the vault held 184 sources, 177 captures and 855 hashed files from six writers.

• The argument. Capture before you write: live pages, down pages and posts change within hours, so the vault came before the page, and captures are only ever added.
• What went wrong. Two collecting agents stopped before committing. Forty sources were recovered from their folders, and two rules follow.
• Figures. The morning in two lanes, from the BBC page, the captures and two git logs; and one vault with many writers, counted from every source.json.

The PHP story's own site, php.oct.2026.newsroom.sgit.ai, on the front page (card 088): a Developing story strip in all three views, fed by php-flaws/build/track.py, which checks the live site. The same five checks are at the top of /php-flaws/, which retires into a pointer when they pass. The briefing pack for the site's team asks for /status.json and for the 40 recovered sources.

One rule widened. admin/build/test_front.mjs checks that the front page's own body links nowhere off the site. It now also allows https://<story>.newsroom.sgit.ai/, a story's own site under the newsroom's name, which this strip exists to link to.