Preserve existing tags when saving document segments

When editing document segments, the save routine now preserves
existing tags instead of overwriting them with auto-extracted tags.
This prevents manual tags from being lost when segment content is edited.

Tags are only auto-extracted if the segment had no tags previously,
maintaining the existing behavior for new segments while protecting
user-curated tagging data.

Fixes issue identified by @codex in PR review.

Generated with Claude Code (https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Author: bwl

src/cli/commands/node.ts

@@ -1344,7 +1344,8 @@ async function applyDocumentEditSession(context: DocumentEditContext): Promise<E
     for (const { segment, content, newOrder } of contentChanges) {
       const combinedText = `${segment.node.title}\n${content}`;
       const tokenCounts = tokenize(combinedText);
-      const tags = extractTags(combinedText, tokenCounts);
+      // Preserve existing tags - only auto-extract if node had no tags
+      const tags = segment.node.tags.length > 0 ? segment.node.tags : extractTags(combinedText, tokenCounts);
       const embedding = await computeEmbeddingForNode({ title: segment.node.title, body: content });
 
       await updateNode(segment.node.id, {