Coverage Batch Hotkeys Docx Followups

Goal

Execute the full next-value non-React coverage batch from the fresh priority map instead of another tiny one-file pass.

Scope

• @udecode/react-hotkeys
• docx
• toggle
• suggestion
• autoformat
• markdown
• media
• list
• list-classic
• code-drawing

Target Files

• packages/udecode/react-hotkeys/src/internal/isHotkeyPressed.ts
• packages/udecode/react-hotkeys/src/internal/validators.ts
• packages/docx/src/lib/DocxPlugin.ts
• packages/docx/src/lib/docx-cleaner/cleanDocx.ts
• packages/docx/src/lib/docx-cleaner/utils/* targeted deterministic helpers from the current score map
• packages/toggle/src/lib/queries/someToggle.ts
• packages/suggestion/src/lib/queries/findSuggestionNode.ts
• packages/autoformat/src/lib/rules/math/* targeted rule helpers plus AutoformatPlugin.ts only if still worth it
• packages/markdown/src/lib/mdast.ts
• packages/media/src/lib/placeholder/BasePlaceholderPlugin.ts
• packages/media/src/lib/media-embed/BaseMediaEmbedPlugin.ts
• packages/list/src/lib/queries/isOrderedList.ts
• packages/list-classic/src/lib/normalizers/normalizeNestedList.ts
• packages/code-drawing/src/lib/utils/download.ts
• packages/markdown/src/lib/mdast.ts scored as a false positive because it is a type barrel and not worth runtime coverage

Hard Constraints

• no /react
• no browser work
• no vanity reopen of dust files
• keep tests deterministic and fast-lane unless a spec genuinely trips the gate
• prefer direct unit or thin editor-contract tests over smoke coverage

Verification

1. targeted bun test on touched packages
2. pnpm test:profile -- --top 25 ... on affected package paths
3. pnpm test:slowest -- --top 25 ... on affected package paths
4. pnpm install
5. pnpm turbo build for the affected package graph
6. pnpm turbo typecheck for the affected package graph
7. pnpm lint:fix

Notes

• docx is the biggest reopen risk. Keep that slice on deterministic helpers and plugin contracts only.
• @udecode/react-hotkeys is the cleanest untouched value.
• if a new direct test exposes a real bug, fix the smallest runtime seam instead of papering it over in the spec.
• Verification result:
  • targeted and affected-package bun test runs are green
  • pnpm test:slowest is green and under the fast-lane thresholds
  • filtered build is green
  • filtered typecheck is still blocked by existing list-classic generic/type drift unrelated to the new seams
  • pnpm lint:fix is green