docs(M-102): fix findings-index false-positive, release <version> bug, land parked site cleanup (#53)

* docs(M-102): fix findings-index false-positive, release <version> bug, land parked site cleanup

Docs triage pass (docs/dev/notes/PROMPT-docs-triage-session.md).

- gen_findings_index.py: D194's row was silently bucketed FIXED (an open
  design item) because it had no distinct status cell, so the generator's
  fallback keyword search matched "fixed" somewhere inside 6,500 chars of
  narrative -- the same misleading-index failure mode that cost a session
  on D227 (see docs/dev/findings.md SS H "D227 -- M-98"/"M-99"). Gave D194
  a real status cell. Also extended LEAD_RE/bucket() with 5 verdict tokens
  already in use in the doc (ANALYZED, GUARDED, DIAGNOSTIC SHIPPED,
  ROOT-CAUSED, NOT A BUG) and added a leading verdict word to 4 more rows
  that had a status cell but no recognized token -- fixes all 10 rows the
  generator previously bucketed "?" (D175 D174 D173 D220 D160 D133 D136
  D156 D153, plus D194). findings-index.csv regenerated; --check passes;
  diff reviewed, only these 10 rows moved.
- ci.yml: release-notes.md's <version> placeholders were never substituted
  before `gh release create --notes-file`, so every published release body
  (including the live v0.1.0 release) ships the literal text. Now sed'd to
  the tag first.
- Landed docs/dev/notes/UNLANDED-docs-site-cleanup-steps3-5.{md,patch}
  (parked M-101, never pushed, branch deleted): deletes the orphaned
  docs/StructureGuide.md (0 tracked references), adds title/description
  frontmatter to 5 published Pages docs, rewrites docs/index.md's stale
  "Phase 2 of 4 / no audio yet / 19-of-21" status text to match the
  current README. The 2 files whose hunks no longer applied via `git am`
  (docs/index.md, docs/dev/agentic-development.md) were redone by hand
  against current main.
- docs/BACKLOG.md (local, gitignored): retargeted a dead
  docs/PORT-LEARNINGS.md reference to docs/porting-notes.md.

Larger triage items (findings.md monster-row convention, full
status-accuracy sweep, duplicate-defect clustering, dev-notes plan
triage, HANDOFF rotation policy) scoped but not started -- see
docs/dev/notes/FOLLOWUP-docs-triage-session.md (gitignored).

Claude-Session: https://claude.ai/code/session_016TaKy6AAsVUTfjkhrX6CjG

* docs/index.md: fix stale in-level-music claim before merge

This PR predates the M-102c/M-105 correction (already landed in the
README via PR #55): in-level music is NOT absent, it plays with some
tracks having wrong-sounding instruments / occasional garbling (D77
fixed, D230 filed for the residual). This branch's own rewritten Status
paragraph still said music 'doesn't yet' play -- fixing it to match
before merge so main doesn't briefly regress to a stale claim this
session already corrected elsewhere.

docs/internals.md

@@ -1,3 +1,8 @@
+---
+title: Internals
+description: Architecture behind the PC port — the software RSP-emulation approach, GoldenEye-vs-Perfect-Dark engine differences, and the phased development plan.
+---
+
 # GoldenEye 007 — PC Port: architecture & plan
 
 > This began as the pre-implementation research note and is kept as the