style: remove em/en dashes repo-wide, add CLAUDE.md rule against them (#890)

Em dashes (—) and en dashes (–) had spread across comments, docs, tests,
and a few UI strings, reading as AI-generated boilerplate rather than
house style. Replaced each with punctuation matching its context: colon
for explanatory clauses, comma for asides, plain hyphen for numeric/legal
ranges (e.g. "21-23§"), "to"/"till" for date ranges, parentheses for
paired-dash asides. messages/en.json and messages/sv.json were fixed by
hand together to keep sv/en in sync.

Left untouched where the dash is the functional subject rather than
decorative punctuation: date-range-parser.ts's separator regex,
charset-repair.ts's CP1252 byte-mapping table (and its test), the SIE
encoding mojibake docs, generic-csv.ts's minus-sign normalizer, the
agent system-prompt files that already instruct against em dashes, and
a golden iXBRL test fixture compared byte-for-byte.

Also fixes two bugs surfaced along the way: an off-by-one in
ApiKeysPanel's scope-label split (a leftover from an earlier partial
pass), and a charset-repair test that had lost the literal en-dash it
exists to verify.

Regenerated the agent atom seed migration (skills:generate) since 27
SKILL.md files changed. Added a CLAUDE.md rule against em/en dashes,
with an explicit carve-out for the functional-dash cases above.

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>

lib/import/shared/encoding.ts

@@ -92,7 +92,7 @@ export function prepareContent(content: string): string {
  * --- U+FFFD heuristic recovery ---
  *
  * When Windows-1252 / Latin-1 bytes are decoded as UTF-8 with `fatal: false`,
- * invalid sequences silently become U+FFFD. The original byte is lost — but
+ * invalid sequences silently become U+FFFD. The original byte is lost, but
  * for Swedish text we can guess from context: the missing letter is almost
  * always one of Å/Ä/Ö (uppercase context) or å/ä/ö (lowercase context).
  *
@@ -133,10 +133,10 @@ const SWEDISH_STEMS = new Set<string>([
   'återbetalning', 'utlägg', 'utgift', 'avdrag',
   // omvänd moms etc.
   'omvänd', 'omvänt', 'omvända',
-  // 3+ letter prepositions/adverbs (skip 2-letter ones — too ambiguous)
+  // 3+ letter prepositions/adverbs (skip 2-letter ones: too ambiguous)
   'från', 'över', 'när', 'där', 'även', 'någon', 'något', 'många',
   'själv', 'små', 'väg', 'gång', 'räkning', 'räntor', 'är',
-  // 'på' — short but extremely common; include explicitly
+  // 'på': short but extremely common; include explicitly
   'på',
   // cities
   'göteborg', 'malmö', 'örebro', 'östersund', 'jönköping', 'linköping',
@@ -190,7 +190,7 @@ function chooseCase(word: string): 'upper' | 'lower' {
 /**
  * Try every Swedish-vowel substitution for the U+FFFDs in `word`, then return
  * the highest-scoring candidate. Returns null if no candidate matches a
- * dictionary stem (i.e. ambiguous — operator must review).
+ * dictionary stem (i.e. ambiguous: operator must review).
  */
 export function recoverWordWithFFFD(word: string): string | null {
   if (!word.includes(REPLACEMENT)) return word
@@ -201,8 +201,7 @@ export function recoverWordWithFFFD(word: string): string | null {
     if (word[i] === REPLACEMENT) positions.push(i)
   }
 
-  // Words with more than ~6 lost bytes blow up the combinatorial space —
-  // bail out rather than spend cycles on something that's likely garbage anyway.
+  // Words with more than ~6 lost bytes blow up the combinatorial space:   // bail out rather than spend cycles on something that's likely garbage anyway.
   const totalCombos = Math.pow(vowels.length, positions.length)
   if (totalCombos > 729) return null