docs(tasks): archive 42 completed board tasks pre-compact

Snapshot resolved tasks (foundation/distro, Ardour/boundaries, Judge UI, triangle
DRY, fleet colour, the sample-grounding fun-sprint, and the dataviz epic #64) so
the active board stays lean. Active board now = pending/in-progress only.

armada/tasks/board-archive.md

+# L'Armada — completed board archive
+
+Snapshot of resolved Claude-Code board tasks, archived before /compact so the active
+board stays lean. Narrative achievement logs live in the numbered `NNN-*.md` files; this
+is just the tracker history (id · subject). Last archived: 2026-06-07.
+
+## Foundation & distribution
+- #1 Scaffold L'Armada toolbox in Tidal repo
+- #7 Settle distribution strategy + RouteNote reality check
+- #9 Seed escales ledger + flotille network graph
+- #14 Build distribution decision SPA (le Sextant)
+
+## Ardour stems · boundaries · per-take EDA
+- #13 Parse Ardour 'Tidal Multi' session → stem map
+- #25 Per-orbit timbral boundary detection for set-splitting
+- #39 v2 boundary bleed: spectral/per-sound novelty (shared-orbit leaks)
+- #54 build_eda.py: per-take EDA emitter (eda_{take}.json) + TakeEDA model
+- #55 EDA-status ledger (Stage 0): map local stems → takes, classify reachability
+- #56 Drop empty/sketch take-types from candidate links (reconciler refinement)
+
+## Judge UI
+- #28 Judge UI: waveform + seekbar + per-moment notes
+- #40 Fix Judge UI orbit annotations: derive from .tidal, not hardcoded ROLE_MAP
+- #41 Judge UI QOL: waveform zoom/scroll + loop-by-select
+
+## Triangle reconciler · catalog DRY · toolbox
+- #42 DRY data layer: pydantic models → JSON Schema → generated TS types
+- #46 Triangle reconciler: ground A↔B↔C into one CatalogView
+- #47 Model CatalogView in models.py (DRY) + regen TS
+- #48 Triangle viz: lit-cell catalog confirmation dashboard
+- #49 pytest suite: UT (parser, agree) + IT-on-real-data + DRY contract
+- #50 tide.py build driver: regenerate all downstream artifacts (parsers-over-copy)
+- #51 Demote catalog.yaml → authored overlay + generated artifact
+- #52 Triangle drawer: ground-truth validators (source + audio + gig)
+- #57 Triangle viz usability pass (PLN feedback)
+- #58 Proper toolbox: load-once + robust serving (no path fragility)
+- #59 Pattern n-grams: duplicated-section detection + 'name each pattern' clustering
+- #60 Surface pattern registry in the triangle drawer + a clusters view
+- #67 Normalize style labels (SSOT in models.py)
+
+## Fleet color language
+- #61 Fleet color language: ontology-in-models → generated tokens
+
+## Sample grounding · feature stack · semantic L2 · dataviz  (the fun-sprint, 2026-06-07)
+- #68 Improve sample-family classifier coverage
+- #69 Sample-family grounding via CLAP zero-shot (sample_classify.py)
+- #72 Featurize: one-shot librosa extractor (sample_features.py)
+- #73 Featurize: TDD UT suite on real samples (kick/bass/hat/break invariants)
+- #75 Featurize: run across corpus samples → sample_features.json
+- #77 feature_eda: correlation matrix + redundancy pruning
+- #78 feature_eda: PCA + intrinsic dimensionality
+- #79 feature_eda: clustering vs CLAP/name/resolver + feature importance
+- #81 Superfeature shortlist → 'ParVagues Unwrapped' exploratory viz  (log 014)
+- #82 Semantic L2: CLAP vibe-search + find-similar (shipped similarity, not tags)  (log 015)
+- #83 Step 1: re-run resolver as clap-fine → land sample_families.json
+- #86 serve.py /vibe + /similar endpoint (CLAP query)
+- #87 Vibe-search + find-similar + vibe-map in unwrapped.html
+- #88 Landing index + track-style tag cloud
+- #64 EPIC: 'Information is Beautiful' dataviz of the corpus — shipped (corpus.html
+  narrative + Unwrapped explorable w/ vibe-search, density field, zoom + landing)