feat(test): add manual spec retriever test script

Introduce a new manual test file `test/manual/test-spec-retriever.ts` that demonstrates how to use the `spec_retriever` agent tool with a query. The script loads embeddings, connects to the LanceDB storage, creates the tool, executes it, and prints the result. This provides a runnable example for developers.

test/manual/test-spec-retriever.ts

@@ -0,0 +1,35 @@
+/**
+ * Manual test script for spec_retriever agent tool
+ * Usage: bun run test/manual/test-spec-retriever.ts "your search query"
+ */
+
+import * as lancedbSdk from "@lancedb/lancedb";
+import { OllamaEmbeddings } from "@langchain/ollama";
+import { createSpecRetrieverTool } from "../../agent_tools";
+import { EMBEDDING_MODEL, STORAGE_DIR } from "../../constants";
+
+async function main() {
+  const query = process.argv[2] || "array.[[DefineOwnProperty]]";
+
+  console.log(`Testing spec_retriever with query: "${query}"`);
+  console.log("Loading database and embeddings...\n");
+
+  const embeddings = new OllamaEmbeddings({
+    model: EMBEDDING_MODEL,
+  });
+
+  const db = await lancedbSdk.connect(STORAGE_DIR);
+  const table = await db.openTable("spec_vectors");
+
+  console.log("Creating tool...\n");
+  const specRetrieverTool = createSpecRetrieverTool(table, embeddings);
+
+  console.log("Executing tool...\n");
+  const result = await specRetrieverTool.func({ query });
+
+  console.log("=== TOOL OUTPUT ===");
+  console.log(result);
+  console.log("\n=== END OUTPUT ===");
+}
+
+main().catch(console.error);