perf: Optimize SQL queries and reduce allocations in hot paths

Change detection queries (embedding/change_detector.rs):
- Replace triple-EXISTS subquery pattern with LEFT JOIN + NULL check
- SQLite now scans embedding_metadata once instead of three times
- Semantically identical: returns docs needing embedding when no
  embedding exists, hash changed, or config mismatch

Count queries (cli/commands/count.rs):
- Consolidate 3 separate COUNT queries for issues into single query
  using conditional aggregation (CASE WHEN state = 'x' THEN 1)
- Same optimization for MRs: 5 queries reduced to 1

Search filter queries (search/filters.rs):
- Replace N separate EXISTS clauses for label filtering with single
  IN() clause with COUNT/GROUP BY HAVING pattern
- For multi-label AND queries, this reduces N subqueries to 1

FTS tokenization (search/fts.rs):
- Replace collect-into-Vec-then-join pattern with direct String building
- Pre-allocate capacity hint for result string

Discussion truncation (documents/truncation.rs):
- Calculate total length without allocating concatenated string first
- Only allocate full string when we know it fits within limit

Embedding pipeline (embedding/pipeline.rs):
- Add Vec::with_capacity hints for chunk work and cleared_docs hashset
- Reduces reallocations during embedding batch processing

Backoff calculation (core/backoff.rs):
- Replace unchecked addition with saturating_add to prevent overflow
- Add test case verifying overflow protection

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

src/embedding/change_detector.rs

@@ -16,30 +16,25 @@ pub fn find_pending_documents(
     last_id: i64,
     model_name: &str,
 ) -> Result<Vec<PendingDocument>> {
+    // Optimized query: LEFT JOIN + NULL check replaces triple-EXISTS pattern.
+    // This allows SQLite to scan embedding_metadata once instead of three times.
+    // Semantically identical: returns documents needing (re-)embedding when:
+    // - No embedding exists (em.document_id IS NULL)
+    // - Content hash changed (em.document_hash != d.content_hash)
+    // - Config mismatch (model/dims/chunk_max_bytes)
     let sql = r#"
         SELECT d.id, d.content_text, d.content_hash
         FROM documents d
+        LEFT JOIN embedding_metadata em
+          ON em.document_id = d.id AND em.chunk_index = 0
         WHERE d.id > ?1
           AND (
-            NOT EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-            )
-            OR EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-                  AND em.document_hash != d.content_hash
-            )
-            OR EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-                  AND (
-                    em.chunk_max_bytes IS NULL
-                    OR em.chunk_max_bytes != ?3
-                    OR em.model != ?4
-                    OR em.dims != ?5
-                  )
-            )
+            em.document_id IS NULL
+            OR em.document_hash != d.content_hash
+            OR em.chunk_max_bytes IS NULL
+            OR em.chunk_max_bytes != ?3
+            OR em.model != ?4
+            OR em.dims != ?5
           )
         ORDER BY d.id
         LIMIT ?2
@@ -69,31 +64,19 @@ pub fn find_pending_documents(
 }
 
 pub fn count_pending_documents(conn: &Connection, model_name: &str) -> Result<i64> {
+    // Optimized query: LEFT JOIN + NULL check replaces triple-EXISTS pattern
     let count: i64 = conn.query_row(
         r#"
         SELECT COUNT(*)
         FROM documents d
-        WHERE (
-            NOT EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-            )
-            OR EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-                  AND em.document_hash != d.content_hash
-            )
-            OR EXISTS (
-                SELECT 1 FROM embedding_metadata em
-                WHERE em.document_id = d.id AND em.chunk_index = 0
-                  AND (
-                    em.chunk_max_bytes IS NULL
-                    OR em.chunk_max_bytes != ?1
-                    OR em.model != ?2
-                    OR em.dims != ?3
-                  )
-            )
-        )
+        LEFT JOIN embedding_metadata em
+          ON em.document_id = d.id AND em.chunk_index = 0
+        WHERE em.document_id IS NULL
+           OR em.document_hash != d.content_hash
+           OR em.chunk_max_bytes IS NULL
+           OR em.chunk_max_bytes != ?1
+           OR em.model != ?2
+           OR em.dims != ?3
         "#,
         rusqlite::params![CHUNK_MAX_BYTES as i64, model_name, EXPECTED_DIMS as i64],
         |row| row.get(0),

src/embedding/pipeline.rs

@@ -103,7 +103,7 @@ async fn embed_page(
     total: usize,
     progress_callback: &Option<Box<dyn Fn(usize, usize)>>,
 ) -> Result<()> {
-    let mut all_chunks: Vec<ChunkWork> = Vec::new();
+    let mut all_chunks: Vec<ChunkWork> = Vec::with_capacity(pending.len() * 3);
     let mut page_normal_docs: usize = 0;
 
     for doc in pending {
@@ -159,7 +159,7 @@ async fn embed_page(
         page_normal_docs += 1;
     }
 
-    let mut cleared_docs: HashSet<i64> = HashSet::new();
+    let mut cleared_docs: HashSet<i64> = HashSet::with_capacity(pending.len());
 
     for batch in all_chunks.chunks(BATCH_SIZE) {
         let texts: Vec<String> = batch.iter().map(|c| c.text.clone()).collect();