Feature: Historical quota/usage snapshots for trend & forecasting

Review the existing quota and usage CSV output and dashboard/reporting flow, along with the read-only pattern in Deploy-QuotaGroups.ps1. The proposed entry point is Get-CapacityTrend.ps1, with snapshots under output/history//. Done means local history is persisted, trends and projections are computed, and the dashboard falls back cleanly when no history exists.

Alignment with toolkit principles

• This stays read-only by default — it does not mutate Azure resources, or (if a write is essential) it follows the opt-in, -WhatIf/ShouldProcess pattern of Deploy-QuotaGroups.ps1.
• It works with Reader access for analysis (writes, if any, document their elevated roles).
• It introduces no telemetry and no data leaving the user's machine.

Problem / motivation

Everything the toolkit produces is point-in-time. The one question it cannot answer is the one
capacity planners most often ask: "How fast is our usage growing, and when will we hit the ceiling?"
There is no notion of history or trend.

Proposed solution

A lightweight, local-only snapshotting layer (no telemetry, no external store):

• Persist each run's quota/usage CSVs into a timestamped local folder (e.g. output/history/<date>/).
• A Get-CapacityTrend.ps1 that loads N snapshots and computes per-family/per-region deltas and a
  simple linear projection to the limit ("at current rate, family X reaches its limit in ~W weeks").
• Dashboard: optional sparkline / trend column when history is present; degrades gracefully to the
  current point-in-time view when it is not.

Stays fully offline and read-only; history lives only on the user's machine.

Alternatives considered

• Push to Log Analytics / a database — rejected: violates the no-telemetry, dependency-light, runs-
  with-Reader principles.

Area

Quota / quota groups, Dashboard / reporting