perf(mcp): Batch semantic code search embeddings to resolve CPU bottleneck and fix linting warnings

server/src/handlers/ast.rs

@@ -2,7 +2,7 @@ use crate::router::McpTool;
 use crate::state::MemoryState;
 use crate::tools::ReadFileSkeletonTool;
 use async_trait::async_trait;
-use serde_json::{json, Value};
+use serde_json::Value;
 use std::sync::Arc;
 use tree_sitter::{Parser, Node};
 

server/src/handlers/workspaces.rs

@@ -596,7 +596,7 @@ impl McpTool for ReadDirectoryArchitectureHandler {
     }
 }
 use crate::tools::SemanticCodeSearchTool;
-use crate::embedding::{generate_embedding_async, cosine_similarity};
+use crate::embedding::{generate_embedding_async, generate_embeddings_async, cosine_similarity};
 
 pub struct SemanticCodeSearchHandler;
 
@@ -620,21 +620,26 @@ impl McpTool for SemanticCodeSearchHandler {
         
         // For MVP, we search across snippets dynamically. A true background codebase indexer would be a separate subsystem.
         let mut results = Vec::new();
+        let mut texts_to_embed = Vec::new();
+        let mut metadata = Vec::new();
         
         let snippets = state.snippets.read_with(|snips| snips.clone());
         for snippet in snippets {
             let combined = format!("{} {} {}", snippet.name, snippet.description, snippet.code);
-            if let Ok(emb) = generate_embedding_async(combined).await {
-                let sim = cosine_similarity(&query_emb, &emb);
-                results.push((sim, snippet.name, snippet.description));
-            }
+            texts_to_embed.push(combined);
+            metadata.push((snippet.name, snippet.description));
         }
         
         let sticky = state.sticky.read_with(|s| s.clone());
         for note in sticky {
-            if let Ok(emb) = generate_embedding_async(note.content.clone()).await {
+            texts_to_embed.push(note.content.clone());
+            metadata.push(("StickyNote".to_string(), note.content.chars().take(200).collect::<String>()));
+        }
+        
+        if let Ok(embeddings) = generate_embeddings_async(texts_to_embed).await {
+            for (emb, meta) in embeddings.into_iter().zip(metadata) {
                 let sim = cosine_similarity(&query_emb, &emb);
-                results.push((sim, "StickyNote".to_string(), note.content.chars().take(200).collect::<String>()));
+                results.push((sim, meta.0, meta.1));
             }
         }