chore(plans): archive the completed playable-hard-offline program

Moves the finished plan and its nine audit and phase reports under
plans/archive/, keeping each program's reasoning next to the plan it
produced. plans/reports/ stays free for active work.

The audits are kept rather than discarded because they hold the baseline
measurements a future change would need to compare against, notably that
removing every guard once changed no level's optimal solution.

plans/archive/260725-2208-playable-hard-offline/plan.md

@@ -19,7 +19,7 @@ All 11 playable levels are degenerate: deleting every guard changes BFS optimal
 moves on every level. Not an authoring problem — four engine defects mean the mechanics never execute.
 CI passed because `levels.solvability.test.js:42-48` asserts only *solvable* and *≤ par*, with no lower bound.
 
-Source reports in `plans/reports/`:
+Source reports in `./reports/` (archived alongside this plan):
 - `puzzle-design-difficulty-audit-260725-2150-level-quality-report.md`
 - `engine-correctness-review-260725-2150-solver-sync-and-state-report.md`
 - `playability-and-ux-audit-260725-2150-player-experience-report.md`

plans/archive/README.md

@@ -0,0 +1,20 @@
+# Archived plans
+
+Completed work programs. Each directory holds the plan and the reports produced
+while executing it, so the reasoning behind a change stays with the change.
+
+Active plans live one level up in `plans/`; new reports go to `plans/reports/`.
+
+| Plan | Date | Outcome |
+|---|---|---|
+| [260725-2208-playable-hard-offline](260725-2208-playable-hard-offline/plan.md) | 2026-07-25/26 | Made the stealth mechanics actually execute, retuned levels 1–11, added offline play, responsive UI, and a hint system. |
+
+## Why the reports are kept
+
+The audits record measurements that are expensive to reproduce and hard to
+believe without evidence — most importantly that, before this work, deleting
+every guard from every level changed the optimal solution by zero moves, and
+the final playable level could be beaten by walking down the left edge and
+across the bottom. If a future change makes levels feel flat again, the
+ablation numbers in `puzzle-design-difficulty-audit-*` and
+`from-level-retune-phase3-*` are the baseline to compare against.