Files
website/server/swagger
wtclaude fff14848f1 feat(moderation): member-raised abuse reports, to site staff only
TEAMS.md §5.6. **Core has had no user-facing report flow of any kind** — the
`moderation`, `mod_notes` and `appeals` tables are all either staff-initiated or
Discord-sanction-shaped, and nothing anywhere let a member say "this is a
problem". That was survivable while every piece of content on the site came from
staff; phase 5 lets players write to each other, so it stops being.

The gap has a specific shape: leaders moderate their own Team's forum, and a
Team's leaders are exactly the people who will not report their own Team. So the
whole point of this queue is a path that routes AROUND a Team's own leadership.
Org lead settled it on 2026-08-18: **reports are site administration only** —
there is no leader-facing view of this queue, not even a read-only one scoped to
their own Team. §5.6's "a leader may also see and act on reports for their own
Team" is not implemented and is not deferred.

`content_reports` is deliberately generic — `target_type` is a VARCHAR so a wiki
page or a news comment becomes a value rather than a table — and the queue is
mounted beside appeals under /admin/moderation rather than under Teams, because a
staffer working a queue should have one place to work.

**§5.6's literal unique key has a defect and this does not copy it.** Written as
(target_type, target_id, reporter_user_id, status) it makes CLOSED rows collide
with each other too: reporter reports a post, staff dismiss it, the behaviour
recurs, they report again — and the second dismissal is an UPDATE into a tuple
that already exists, so working the queue starts throwing duplicate-key errors on
the first repeat reporter. The key is on a generated `open_marker` instead, the
same trick `team_forum_grants.active_marker` uses: 1 while open, NULL once
closed, and NULLs are distinct — which is what §5.6's prose asks for, "one open
report per (target, reporter)".

Two other departures from the doc, both small and both flagged in the docs PR:
`handled_note`, because a queue whose resolution reason lives only in an
activity_log line is one where the next staffer to see a repeat report cannot
find out why the last was dismissed; and a CASCADE on `team_id`, so a deleted
Team does not leave a queue full of reports about content that no longer exists.

Also here: a report is filed against a target the model verifies really belongs to
the Team the request came through, or the queue's per-Team filter would quietly be
lying; the queue resolves every row's target in three batched reads rather than
N+1, which is §5.6's fourth rule (uploader, size and sniffed type without
hunting) actually paying for §5.5.4's attribution table; a target that has since
been hard-deleted comes back null and the report still lists, because "somebody
reported this and by the time we looked it was gone" is a fact a moderator needs;
and every transition writes activity_log, `dismissed` included — a queue where
acting is audited and declining to act is not is one where the cheapest way to
make a report vanish leaves no trace.

`teams_forum_edit_window_minutes` gains its range validation on the admin settings
PUT and is seeded at 15, so the value on the settings screen is the value in
force. Route manifest and OpenAPI regenerated: 6 operations added, 0 lost.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-08-18 12:51:42 -05:00
..