fix(ui): validate PowerUse window, explain effort and Prime Claude in Settings

Three usability-audit findings in Settings:
- An unparseable PowerUse window (e.g. "9:0") used to save silently as null
  (= round the clock), the opposite of what the typo implied. Validate() now
  rejects an unparseable start/end unless blank; the worker-side read-time
  tolerance is unchanged.
- "Effort" column header gets a tooltip explaining it's the model's reasoning
  effort (cost/depth tradeoff).
- The "Prime Claude" tab gets a tooltip explaining scheduled Claude runs
  (Ping / Fill My Day / custom prompt).

New locale keys added to en+de. Updated the PowerUse round-trip test that
asserted the old silent-normalize behavior.

src/ClaudeDo.Localization/locales/en.json

@@ -11,6 +11,7 @@
       "categoryGeneral": "General",
       "categoryExecution": "Execution",
       "categoryPrime": "Prime Claude",
+      "categoryPrimeHint": "Scheduled Claude runs: Ping, Fill My Day, or a custom prompt, at fixed times.",
       "categoryWorktrees": "Worktrees",
       "categoryFiles": "Files & Prompts",
       "categorySkills": "Session Skills",
@@ -27,6 +28,7 @@
       "model": "Model",
       "maxTurns": "Max turns",
       "effort": "Effort",
+      "effortHint": "Reasoning effort the model applies. Higher costs more and reasons deeper.",
       "modelPresets": "Per-model defaults",
       "modelPresetsHint": "Reasoning effort and turn budget used when a session runs under that model. List- and task-level max-turn overrides still win.",
       "maxTurnsCeiling": "Max turns ceiling",
@@ -50,6 +52,7 @@
       "powerUseWindowPlaceholder": "HH:mm",
       "powerUseHandoffPct": "Hand off at (% of 5h)",
       "powerUseWindowHint": "Leave both times empty for round the clock. A window is checked whenever a handler phase starts, so a chain that runs past the end time continues without PowerUse.",
+      "powerUseWindowInvalid": "PowerUse window times must be HH:mm, or left blank.",
       "throttleStagesHeading": "Throttle stages",
       "throttleStagesHint": "Parallelism is capped before the hard gate: soft caps at 2 slots, hard at 1. Edited by dragging the markers on the usage-monitor gauges. Opening the monitor closes these settings — save first if you changed anything.",
       "throttleStagesValues": "5h: soft {0}% / hard {1}%  ·  7d: soft {2}% / hard {3}%",