@zabaca/lattice 0.1.2 → 0.2.0

package/README.md

@@ -108,13 +108,12 @@ lattice status --verbose        # Include detailed change information
 
 ### `lattice search`
 
-Search the knowledge graph.
+Semantic search across the knowledge graph.
 
 ```bash
-lattice search --semantic "query"    # Semantic search using embeddings
-lattice search --keyword "term"      # Keyword-based search
-lattice search --type concept        # Filter by entity type
-lattice search --limit 10            # Limit results
+lattice search "query"                    # Search all entity types
+lattice search --label Technology "query" # Filter by entity label
+lattice search --limit 10 "query"         # Limit results (default: 20)
 ```
 
 ### `lattice stats`

package/dist/cli.js

@@ -2385,98 +2385,65 @@ Relationship Types (${stats.relationshipTypes.length}):`);
       process.exit(1);
     }
   });
-  program.command("search").description("Search for nodes in the graph or perform semantic search").option("-l, --label <label>", "Filter by node label").option("-n, --name <name>", "Filter by name (substring match)").option("-s, --semantic <query>", "Perform semantic/vector search on documents").option("--limit <n>", "Limit results", "20").action(async (options) => {
+  program.command("search <query>").description("Semantic search across the knowledge graph").option("-l, --label <label>", "Filter by entity label (e.g., Technology, Concept, Document)").option("--limit <n>", "Limit results", "20").action(async (query, options) => {
     let app;
     try {
       app = await NestFactory3.createApplicationContext(AppModule, {
         logger: ["error"]
       });
       const graph = app.get(GraphService);
-      if (options.semantic) {
-        const embedding = app.get(EmbeddingService);
-        const limit2 = Math.min(parseInt(options.limit, 10), 100);
-        try {
-          const queryEmbedding = await embedding.generateEmbedding(options.semantic);
-          const results2 = await graph.vectorSearchAll(queryEmbedding, limit2);
-          console.log(`
-=== Semantic Search Results for "${options.semantic}" ===
-`);
-          if (results2.length === 0) {
-            console.log(`No results found with semantic search.
-`);
-            await app.close();
-            process.exit(0);
-          }
-          results2.forEach((result2, idx) => {
-            console.log(`${idx + 1}. [${result2.label}] ${result2.name}`);
-            if (result2.title) {
-              console.log(`   Title: ${result2.title}`);
-            }
-            if (result2.description && result2.label !== "Document") {
-              const desc = result2.description.length > 80 ? result2.description.slice(0, 80) + "..." : result2.description;
-              console.log(`   ${desc}`);
-            }
-            console.log(`   Similarity: ${(result2.score * 100).toFixed(2)}%`);
-          });
-          console.log();
-          await app.close();
-          process.exit(0);
-        } catch (semanticError) {
-          const errorMsg = semanticError instanceof Error ? semanticError.message : String(semanticError);
-          console.error("Semantic search error:", errorMsg);
-          if (errorMsg.includes("no embeddings") || errorMsg.includes("vector")) {
-            console.log(`
-Note: Semantic search requires embeddings to be generated first.`);
-            console.log(`Run 'lattice sync' to generate embeddings for documents.
-`);
-          }
-          await app.close();
-          process.exit(1);
-        }
-      }
-      let cypher;
+      const embedding = app.get(EmbeddingService);
       const limit = Math.min(parseInt(options.limit, 10), 100);
-      if (options.label && options.name) {
-        const escapedLabel = options.label.replace(/`/g, "\\`");
-        const escapedName = options.name.replace(/'/g, "\\'");
-        cypher = `MATCH (n:\`${escapedLabel}\`) WHERE n.name CONTAINS '${escapedName}' RETURN n LIMIT ${limit}`;
-      } else if (options.label) {
-        const escapedLabel = options.label.replace(/`/g, "\\`");
-        cypher = `MATCH (n:\`${escapedLabel}\`) RETURN n LIMIT ${limit}`;
-      } else if (options.name) {
-        const escapedName = options.name.replace(/'/g, "\\'");
-        cypher = `MATCH (n) WHERE n.name CONTAINS '${escapedName}' RETURN n LIMIT ${limit}`;
+      const queryEmbedding = await embedding.generateEmbedding(query);
+      let results;
+      if (options.label) {
+        const labelResults = await graph.vectorSearch(options.label, queryEmbedding, limit);
+        results = labelResults.map((r) => ({
+          name: r.name,
+          label: options.label,
+          title: r.title,
+          score: r.score
+        }));
       } else {
-        cypher = `MATCH (n) RETURN n LIMIT ${limit}`;
+        results = await graph.vectorSearchAll(queryEmbedding, limit);
       }
-      const result = await graph.query(cypher);
-      const results = result.resultSet || [];
+      const labelSuffix = options.label ? ` (${options.label})` : "";
       console.log(`
-=== Search Results (${results.length} nodes) ===
+=== Semantic Search Results for "${query}"${labelSuffix} ===
 `);
       if (results.length === 0) {
-        console.log(`No nodes found matching criteria.
+        console.log(`No results found.
+`);
+        if (options.label) {
+          console.log(`Tip: Try without --label to search all entity types.
 `);
+        }
         await app.close();
         process.exit(0);
       }
-      results.forEach((row) => {
-        const node = row[0];
-        const labels = (node.labels || []).join(", ");
-        const name = node.properties?.name || "unnamed";
-        console.log(`[${labels}] ${name}`);
-        if (node.properties?.description) {
-          console.log(`  Description: ${node.properties.description}`);
+      results.forEach((result, idx) => {
+        console.log(`${idx + 1}. [${result.label}] ${result.name}`);
+        if (result.title) {
+          console.log(`   Title: ${result.title}`);
         }
-        if (node.properties?.importance) {
-          console.log(`  Importance: ${node.properties.importance}`);
+        if (result.description && result.label !== "Document") {
+          const desc = result.description.length > 80 ? result.description.slice(0, 80) + "..." : result.description;
+          console.log(`   ${desc}`);
         }
+        console.log(`   Similarity: ${(result.score * 100).toFixed(2)}%`);
       });
       console.log();
       await app.close();
       process.exit(0);
     } catch (error) {
-      console.error("Error:", error instanceof Error ? error.message : String(error));
+      const errorMsg = error instanceof Error ? error.message : String(error);
+      console.error("Error:", errorMsg);
+      if (errorMsg.includes("no embeddings") || errorMsg.includes("vector")) {
+        console.log(`
+Note: Semantic search requires embeddings to be generated first.`);
+        console.log(`Run 'lattice sync' to generate embeddings for documents.
+`);
+      }
       if (app)
         await app.close();
       process.exit(1);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@zabaca/lattice",
-  "version": "0.1.2",
+  "version": "0.2.0",
   "description": "Human-initiated, AI-powered knowledge graph for markdown documentation",
   "type": "module",
   "bin": {