Merge train 4843-8 (CI only; not for merge)

[erikdarlingdata]

Merge train, not for merge. This PR exists only to run CI once. Its branch is train 7b (#5153, not landed yet) plus the squash of each PR below, in this order, so CI here covers dev + train 7b + train 8. After train 7b lands, each PR below is squash-merged by itself in the same order, and dev's final tree is checked against this branch's tree. Then this PR is closed unmerged.

Part of #4966 and #4843.

PRs in this train, in order

• MCP: Darling's four data-start twins, and Lite's config change, collection log, plan correction and memory pressure tools, say where their data starts (part of Say where the data starts on every time-ranged tab, in Lite and both Darling viewers #4966) (MCP: Darling's four data-start twins, and Lite's config change, collection log, plan correction and memory pressure tools, say where their data starts (part of #4966) #5147)
• Web grids: column groups with a toggle strip, first used on Memory Conditions and Node OOM (Web grids: column groups with a toggle strip, first used on Memory Conditions and Node OOM #5152)
• The web Alert History page gains Dismiss Selected and Dismiss All (The web Alert History page gains Dismiss Selected and Dismiss All #5151)

Train head 890131e9c, tree 24ce4a38a, stacked on train 7b 0fc276c33. A local merged-tree check ran 1,584 tests, including the whole guard family. The one failure is a class that cannot load on macOS; CI runs it.

[erikdarlingdata]

Train 8 landed: each PR above was squash-merged by itself in order, and dev's tree after the last one (7334986) is byte-identical to this PR's tested tree. Closing unmerged, as planned.