From 662e59067c9500a776e3758b04892516d190feae Mon Sep 17 00:00:00 2001
From: Fuwn <contact@fuwn.me>
Date: Tue, 3 Feb 2026 21:46:27 -0800
Subject: feat(mcp): Add text-based semantic search tool

---
 packages/mcp/src/index.ts | 47 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/packages/mcp/src/index.ts b/packages/mcp/src/index.ts
index 46aa487..4ab8af8 100644
--- a/packages/mcp/src/index.ts
+++ b/packages/mcp/src/index.ts
@@ -567,6 +567,53 @@ server.tool(
 		};
 	},
 );
+server.tool(
+	"search_memories_by_text",
+	"Search memories using text query with auto-generated embedding",
+	{
+		query: z.string().describe("The search query text"),
+		threshold: z
+			.number()
+			.optional()
+			.describe("Minimum similarity threshold (0-1, default 0.7)"),
+		limit: z
+			.number()
+			.optional()
+			.describe("Maximum number of results (default 10)"),
+		projectId: z.string().optional().describe("Filter by project ID"),
+		folderId: z.string().optional().describe("Filter by folder ID"),
+	},
+	async (parameters) => {
+		if (!embeddingService) {
+			return {
+				content: [
+					{
+						type: "text" as const,
+						text: "Embedding service not configured. Set OPENAI_API_KEY or IMEMIO_EMBEDDING_TYPE=local.",
+					},
+				],
+				isError: true,
+			};
+		}
+
+		const queryEmbedding = await embeddingService.generate(parameters.query);
+		const searchResults = await memoryStore.search(queryEmbedding, {
+			threshold: parameters.threshold,
+			limit: parameters.limit,
+			projectId: parameters.projectId,
+			folderId: parameters.folderId,
+		});
+
+		return {
+			content: [
+				{
+					type: "text" as const,
+					text: JSON.stringify(searchResults, null, 2),
+				},
+			],
+		};
+	},
+);
 server.tool(
 	"generate_embedding",
 	"Generate an embedding vector for text content",
-- 
cgit v1.2.3