enhance(#4014): an unreadable directory must not report as an empty one (#4163)

* test(#4014): add failing-first coverage for unreadable-vs-empty directory scope (epic #3473 B4)

* fix(#4014): an unreadable directory must not report as an empty one (epic #3473 B4)

* test(#4014): update hardcoded generateSlugInternal closing-brace line after import shift

src/core-utils.cts's new #4014 import block shifted every subsequent line by
6, moving generateSlugInternal's real closing brace from line 193 to 199.
tests/slug-derivation-drift-guard.test.cjs's MAJOR-1 fixture hardcodes that
line number to plant a synthetic violation immediately after the function's
real body; the guard script itself locates the boundary dynamically via
brace-matching and needed no change.

* docs(#4014): document the unreadable-directory scope signal and add changeset

* docs(#4014): backfill changeset PR number to #4163

* test(#4014): kill pre-existing core-utils.cjs mutation-score gap, unrelated to this issue's diff

---------

Co-authored-by: sim <sim@local>

.changeset/lively-birds-click.md

@@ -0,0 +1,5 @@
+---
+type: Changed
+pr: 4163
+---
+**`roadmap analyze`, `gap-checker`, and `init`'s JSON output now distinguish an unreadable phase directory from a genuinely empty one** — a new `context_scope`/`phase_dir_scope` field (`'complete'` or `'unreadable'`) sits alongside the existing `has_context`/`context_read_error` fields, so a permission or I/O failure reading a phase directory is no longer indistinguishable from a phase that simply has no context file yet. `init manager`'s previously-silent read failure (a bare empty catch) now surfaces the same signal. (#4014)