chore(todo): mark refactoring program nearly complete, scope the deferred harness rewrite

1 files changed, 32 insertions, 52 deletions

diff --git a/todo.org b/todo.org
index b38dcc767..575102256 100644
--- a/todo.org
+++ b/todo.org
@@ -105,59 +105,39 @@ ai-term toggle-off teardown + working-buffer swap (62fee96b); calendar-sync
 per-event exception parser (23f405b4); dirvish playlist-target resolution
 (a1ca2fb0); custom-case per-word title-case decision (4cc9ca0b).
 
-*** Remaining — big single-file (characterization tests first)
-- custom-comments.el: the divider/box render-skeleton duplication — simple-divider vs
-  padded-divider (~40 lines each, differ only in the padding loop) and box vs
-  heavy-box (~45 each, heavy-box just adds two blank inserts); plus the comment-syntax
-  prologue repeated 7x and the doubled-semicolon / length-option constants. Extract a
-  shared emit-prefix helper + a border/text/border emitter parameterized by padding
-  lambda + extra-blank-lines. Insertion-order-sensitive: characterize each generator's
-  output before refactoring. The heaviest item.
-- dwim-shell-config.el: ~46 =cj/dwim-shell-commands-*= defuns are trapped in the
-  =use-package dwim-shell-command :config=, so untestable under =make test=. For the
-  branching ones (=-video-trim= pcase x3, =-text-to-speech= darwin/linux,
-  =-extract-archive=/zip/tar single-vs-multi), extract each command's command-string
-  construction into a top-level pure =cj/dwim-shell--<name>-command= (takes prompted
-  values, returns the template string), leaving a thin interactive wrapper in :config.
-  Mirrors the existing =cj/dwim-shell--dated-backup-command= pattern. Do the
-  high-value branching commands; the trivial ones can stay.
-
-*** Remaining — theme-studio (scripts/theme-studio/)
-Suite: =make check= (Python/Node/ERT) + =./run-tests.sh= (browser gates) +
-=make check-generated= (byte-identical html) + =make coverage=. After ANY
-generate.py-output change, stage theme-studio.html in the SAME commit
-(check-generated compares to the working tree, not HEAD).
-- app-core.js: =dropdownRowTextColor= is exported + has 4 tests but no runtime caller
-  (live path computes inline at app.js:82). Decide: wire it into the dropdown popup
-  row painting, or delete it + its 4 tests. Needs Craig's intent — default to delete.
-- generate.py: ~230 lines of module-level build run at import; =face_coverage.py= does
-  =import generate= just for two constants and pays the whole cost. Wrap the assembly
-  in =build()= gated behind =__main__=; keep UI_FACES/CATS/COLS cheap module
-  constants. (The CRITICAL item.)
-- capture-default-faces.py: =condition_matches= (166-206) has parallel dict-branch and
-  list-branch clause checkers encoding the same four rules + constants twice. Normalize
-  both shapes to one mapping, run one set of checks. NOT in =make check= — verify by
-  running it.
-- face_coverage.py: =bucket_from_source= (118) and =bucket_of_source= (157) duplicate
-  the elpa/user/builtin path-kind detection. Extract =path_kind(path)= and have both
-  map its result to their own vocabulary.
-- browser-gates.js (HIGHEST RISK — rewrites the harness that verifies everything):
-  ~39/44 gates copy-paste the =let ok=true;const notes=[];const A=(c,n)=>{...}= setup +
-  the =document.title=...; result-div= postamble (note format already drifted: 17 use
-  " | " vs 24 " fails="). Extract one =gate(name, body)= helper. CRITICAL CONSTRAINT:
-  each gate's =if(...)= MUST keep the literal substring =location.hash==='#NAMEtest'=
-  because run-tests.sh:76 discovers gates by grepping exactly that — a registry/loop
-  that hides the hash check breaks discovery (silent false-green). Pair with a
-  =withSavedState(keys, body)= helper for the ~13 mutating gates' inconsistent
+*** DONE — big single-file + theme-studio (2026-06-20 afternoon, no-approvals run)
+Both big single-file items shipped: dwim-shell branching command builders
+(f93b4615); custom-comments divider/box generator dedup (42f0c88a). Five of the
+six theme-studio items shipped: face_coverage path_kind (9a52370b),
+capture-default-faces condition_matches unify (28b4d1cf), dropdownRowTextColor
+delete (10a56789), test-file inline-integrity dedup — subTest loop + shared
+inline-strip.mjs (13969c70), generate.py lazy _build()/__getattr__ (6df4ebdc),
+browser-gates assertPreviewFaces for the 3 preview gates (5627f137).
+
+*** Remaining — browser-gates harness rewrite (HIGHEST RISK, deferred for review)
+Two parts of the browser-gates.js item are intentionally NOT done in the
+autonomous no-approvals run — they rewrite the harness that verifies everything,
+so a subtle helper bug manufactures silent false-greens across all 44 gates:
+- =gate(name, body)= helper for the ~39 gates' =let ok;notes;A=...; title; result-div=
+  boilerplate. LOW value (pure boilerplate; note format drifted 17 " | " vs 24
+  " fails="). CRITICAL CONSTRAINT on any attempt: each gate keeps its literal
+  =location.hash==='#NAMEtest'= (run-tests.sh greps it to discover gates) — wrap only
+  the body, e.g. =if(location.hash==='#x')gate('X','x',A=>{...})=. Verify: all 44 gates
+  green AND a deliberately-broken assertion still FAILS (proved feasible — chrome is
+  available; the assertPreviewFaces commit ran exactly that check).
+- =withSavedState(keys, body)= for the ~13 mutating gates' inconsistent
   PALETTE/MAP/UIMAP/SYNTAX snapshot-restore (7 mutating gates currently restore
-  nothing). Verify: all gates green AND a deliberately-broken assertion still FAILS
-  (prove the harness can't manufacture greens). Also =assertPreviewFaces= for the 3
-  copy-pasted preview-face validators (mdtest/mupreviewtest/gnustest).
-- theme-studio test files: =plan(overrides)= factory for the ~30 planPaletteGenerator
-  full-option-literal calls (test-app-core.mjs + test-palette-generator-core.mjs); one
-  shared =stripExports= (reimplemented 3x in app-core/colormath/app-util tests, must
-  stay aligned with generate.py's strip per test_generate.py:48); and an
-  =assertInlinedVerbatim(name)= loop for the 5 inline-integrity cases.
+  NOTHING — a real state-pollution bug, not just dedup). Needs per-gate key analysis.
+Both warrant Craig's eye before/after given the harness-rewrite risk. The
+=assertPreviewFaces= part of this item is already done (5627f137).
+
+*** Remaining — item-8 plan() factory (deferred, low value)
+The =plan(overrides)= factory for the ~30 planPaletteGenerator calls (test-app-core.mjs
++ test-palette-generator-core.mjs) was deferred. The calls pass heterogeneous options
+(scheme/accentCount/sourceMode/vibe/intent vary per call); a factory only dedups the
+constant spanCount:0/rng and would hide which options each test actually exercises —
+premature abstraction over varying calls. The other two item-8 parts (subTest loop +
+shared stripExports) shipped in 13969c70.
 
 *** WON'T-DO (do not re-attempt — assessed and rejected)
 - theme-studio buildTable/buildUITable/buildPkgTable merge: genuine per-tier divergence