docs/design/docs-wordprocessor-roadmap.md (3)

110-111: Clarify whether superscript and subscript are mutually exclusive.

The data model allows both superscript and subscript to be set simultaneously. In typical word processors, these formatting options are mutually exclusive. Consider adding a note specifying the expected behavior when both are true (e.g., subscript takes precedence, or the combination is invalid).

📝 Proposed clarification

   // Phase 2
   href?: string;
   backgroundColor?: string;
-  superscript?: boolean;
-  subscript?: boolean;
+  superscript?: boolean;  // mutually exclusive with subscript
+  subscript?: boolean;     // mutually exclusive with superscript
 }

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@docs/design/docs-wordprocessor-roadmap.md` around lines 110 - 111, The docs
currently allow both superscript and subscript to be true simultaneously; update
the specification to clarify expected behavior by either (a) declaring them
mutually exclusive and stating that validators or the editor must enforce only
one may be true at a time, or (b) defining a clear precedence rule (e.g.,
subscript overrides superscript when both are set). Reference the superscript
and subscript fields in the text and add a short note about how clients should
handle validation or rendering when both flags are present.

---

139-139: Clarify list-exit behavior for nested levels.

The spec states "Enter on empty list item exits list (converts to paragraph)." Does this apply only to top-level items, or does it outdent one level per Enter until reaching the top? Google Docs outdents progressively; specifying this behavior will guide implementation.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@docs/design/docs-wordprocessor-roadmap.md` at line 139, The spec phrase
"Enter on empty list item exits list (converts to paragraph)" is ambiguous for
nested lists; update the spec text to explicitly state the intended behavior
(either "outdent one level per Enter on an empty list item until reaching
top-level, then convert to paragraph" or "only exits the list when at top-level
and converts to paragraph"), and make this change where that phrase appears in
the document so implementers know whether to implement progressive outdenting
(Google Docs style) or a single-step exit to paragraph.

---

136-136: Specify marker styles for list levels beyond 2.

Line 89 allows listLevel from 0 to 8 (9 levels), but only three bullet styles (●, ○, ■) and three ordered patterns are specified. Clarify what markers are used for levels 3-8. Do the patterns cycle, or does level 2+ reuse the last pattern?

📋 Example clarification

-- Bullet markers: unordered (●, ○, ■ by level), ordered (1. 2. 3., a. b. c., i. ii. iii.)
+- Bullet markers: unordered cycles through ● (level 0), ○ (level 1), ■ (level 2), then repeats; ordered cycles through decimal (1. 2. 3.), lower-alpha (a. b. c.), lower-roman (i. ii. iii.), then repeats

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@docs/design/docs-wordprocessor-roadmap.md` at line 136, The document
currently allows listLevel values 0–8 but only defines three unordered markers
(●, ○, ■) and three ordered patterns (1., a., i.); update the docs to explicitly
state how markers are selected for levels 3–8 by either (a) defining a full
mapping for listLevel 0–8 for both unordered and ordered lists, or (b) declaring
the cycling rule (e.g., cycle back to level 0 markers or repeat the last marker)
and showing the exact sequence; reference the term listLevel and the existing
"Bullet markers" / "ordered patterns" entries and include a small table or list
mapping each listLevel (0..8) to its marker so consumers know whether levels 3–8
reuse or cycle marker styles.