fix: process embeddings sequentially, shorten prompts, fix tests

- main.ts: process files sequentially (not Promise.all) to avoid concurrent
  embedding requests hammering Ollama; batch size reduced to 1
- vectorization.ts: shorten embedding prompts from 1000 to 500 chars,
  limit headings to 5, remove frontmatter from prompt to stay well within
  embedding model context window
- Update vectorization and indexing-pipeline tests for new prompt format

tests/vectorization.test.ts

@@ -200,8 +200,7 @@ describe('ContentVectorizer', () => {
       expect(prompt).toContain('This is the first paragraph');
       expect(prompt).toContain('Main Heading');
       expect(prompt).toContain('Sub Heading');
-      expect(prompt).toContain('test');
-      expect(prompt).toContain('2024-01-01');
+      // Frontmatter is no longer included in embedding prompts
     });
 
     it('should handle empty content fields gracefully', () => {
@@ -221,8 +220,7 @@ describe('ContentVectorizer', () => {
 
       const prompt = (vectorizer as any).createPrompt(chunk);
       expect(prompt).toContain('Only content');
-      // JSON.stringify({}) produces "{}", which is truthy so it's included
-      expect(prompt).toContain('{}');
+      // Frontmatter is no longer included in embedding prompts
     });
 
     it('should limit content length', () => {
@@ -243,7 +241,7 @@ describe('ContentVectorizer', () => {
 
       const prompt = (vectorizer as any).createPrompt(chunk);
       expect(prompt).not.toContain('a'.repeat(1500));
-      expect(prompt).toContain('a'.repeat(1000));
+      expect(prompt).toContain('a'.repeat(500));
     });
 
     it('should handle missing frontmatter gracefully', () => {