Editor: document tone, right-click word lookup, expanded stats

Four enhancements to make the editor fit real school usage:

- Per-document tone (academic/professional/casual/humorous/creative/
  persuasive/general): new documents.tone column (migration 0002), threaded
  through the docs API, a bilingual ToneSelect dropdown on the title row, and
  injected into the grammar-checkpoint LLM prompt so advice fits the register.
  The voice pass stays tone-agnostic.

- Right-click word lookup: a new offline `lexicon` package serves definitions
  (Wordset, modern ESL-friendly glosses) and synonyms (WordNet synsets first,
  then frequency+stopword-ranked Moby for breadth) from gzipped embedded data,
  behind /api/word/{word} with light morphology. The WordCard popover shows the
  definition and tappable synonym pills that swap the word in place.

- Expanded writing stats: clicking the word count opens a StatsPanel with page
  count, sentences, paragraphs, reading time, average word length, word variety,
  and Flesch-Kincaid reading level — all computed client-side.

Claude-Session: https://claude.ai/code/session_016Yr6jELuRc7hyzYLccQKZd

internal/llm/checkpoint.go

@@ -31,9 +31,9 @@ type checkpointResponse struct {
 // RunCheckpoint sends the grammar checkpoint and parses the JSON result. It
 // applies the latency-guard truncation and the checkpoint sampling parameters
 // from the spec.
-func RunCheckpoint(ctx context.Context, client LLMClient, contentText string) ([]RawSuggestion, error) {
+func RunCheckpoint(ctx context.Context, client LLMClient, contentText, tone string) ([]RawSuggestion, error) {
 	raw, err := client.Complete(ctx, CompletionRequest{
-		Messages:          CheckpointMessages(TruncateDoc(contentText)),
+		Messages:          CheckpointMessages(TruncateDoc(contentText), tone),
 		MaxTokens:         1024,
 		Temperature:       0.3,
 		RepetitionPenalty: 1.15,

internal/llm/prompts.go

@@ -8,7 +8,7 @@ const checkpointSystemPrompt = `You are a warm, encouraging writing assistant he
 	`Analyze the text below and identify up to 5 issues: grammar errors, unnatural phrasing, ` +
 	`incorrect idiom usage, or unclear sentences that are common ESL patterns.
 
-Be specific, friendly, and explain WHY each suggestion improves the writing.
+Be specific, friendly, and explain WHY each suggestion improves the writing.%s
 
 Respond ONLY with valid JSON. No preamble, no markdown fences. Format:
 {
@@ -24,11 +24,32 @@ Respond ONLY with valid JSON. No preamble, no markdown fences. Format:
 
 If the writing looks good, return: {"suggestions": []}`
 
+// toneGuidance returns a sentence steering the checkpoint toward the writer's
+// chosen tone, or "" for the neutral default. The clause is appended to the
+// checkpoint instructions so the model's phrasing suggestions fit the target
+// register (e.g. an academic essay vs a casual journal). Unknown values fall
+// back to no steering, so a stray tone string is harmless.
+func toneGuidance(tone string) string {
+	clause, ok := map[string]string{
+		"academic":     "formal, academic, and objective — suited to a school essay or research paper",
+		"professional": "polished and professional — suited to a workplace email or report",
+		"casual":       "relaxed, friendly, and conversational",
+		"humorous":     "light, playful, and good-humored",
+		"creative":     "vivid, expressive, and imaginative — suited to a story or personal narrative",
+		"persuasive":   "confident and persuasive — suited to an argument or opinion piece",
+	}[tone]
+	if !ok {
+		return ""
+	}
+	return "\n\nThe writer wants this document to read as " + clause + ". When phrasing could " +
+		"be improved, prefer suggestions that fit that tone, and gently flag wording that clashes with it."
+}
+
 // CheckpointMessages builds the message array for a grammar checkpoint over the
-// given (already-truncated) document text.
-func CheckpointMessages(contentText string) []Message {
+// given (already-truncated) document text, steered toward the document's tone.
+func CheckpointMessages(contentText, tone string) []Message {
 	return []Message{
-		{Role: "system", Content: checkpointSystemPrompt},
+		{Role: "system", Content: fmt.Sprintf(checkpointSystemPrompt, toneGuidance(tone))},
 		{Role: "user", Content: contentText},
 	}
 }

internal/llm/voice.go

@@ -16,7 +16,10 @@ const VoiceInterval = 20 * time.Second
 // for a Tier-1 voice pass and parses the JSON result. It reuses the checkpoint's
 // tolerant parser and a larger token budget, since one pass may flag several
 // passages. Each flag carries a null replacement (awareness-only).
-func RunVoice(ctx context.Context, client LLMClient, contentText string) ([]RawSuggestion, error) {
+// The tone argument is accepted for a uniform pass signature but ignored: voice
+// consistency is judged against the document's own established voice, not an
+// externally-chosen register.
+func RunVoice(ctx context.Context, client LLMClient, contentText, _ string) ([]RawSuggestion, error) {
 	raw, err := client.Complete(ctx, CompletionRequest{
 		Messages:          VoiceMessages(contentText),
 		MaxTokens:         2048,