@closedloop-ai/mcp-client 1.1.0 → 1.1.2

package/README.md

@@ -1,15 +1,17 @@
 # ClosedLoop MCP Server
 
-A lightweight Model Context Protocol (MCP) server that provides AI clients with access to ClosedLoop customer feedback data.
+A lightweight Model Context Protocol (MCP) server that provides AI clients with access to ClosedLoop customer feedback data and advanced search capabilities.
 
 ## Features
 
 - **MCP Protocol Support**: Full MCP protocol implementation for AI client integration
 - **API Key Authentication**: Secure team-based API key authentication
 - **Team Isolation**: All data access is scoped to the team that owns the API key
-- **Two Core Tools**:
+- **Advanced Search**: Full-text search with relevance ranking across 8+ insight fields
+- **Three Core Tools**:
   - `list_insights`: Get customer insights with date range and pagination
   - `get_insight_detail`: Get detailed information about specific insights
+  - `search_insights`: **NEW** Advanced full-text search with filtering and relevance ranking
 
 ## Installation
 
@@ -60,14 +62,59 @@ Get detailed information about a specific insight item.
 **Parameters:**
 - `insight_id` (required): UUID of the insight item
 
+### search_insights ⭐ NEW
+Advanced full-text search across customer insights with relevance ranking and comprehensive filtering.
+
+**Key Features:**
+- **Full-text search** across 8+ insight fields
+- **Relevance ranking** - most relevant results appear first
+- **Phrase search** and fuzzy matching
+- **Field-specific search** - search specific fields or all fields
+- **Advanced filtering** - category, severity, date range, source
+- **Language agnostic** - works with any language
+
+**Parameters:**
+- `query` (required): Search query text
+- `fields` (optional): Specific fields to search in
+- `category` (optional): Filter by insight category
+- `severity` (optional): Filter by severity level
+- `date_from` (optional): Start date filter
+- `date_to` (optional): End date filter
+- `source` (optional): Filter by data source
+- `page` (optional): Page number (default: 1)
+- `limit` (optional): Items per page (default: 20, max: 100)
+
+**Searchable Fields:**
+- `signal_title` - Insight title
+- `content` - Raw feedback content
+- `pain_point` - Customer pain points
+- `workaround` - Customer workarounds
+- `use_case` - Use case descriptions
+- `feature_area` - Feature categorization
+- `competitor_gap` - Competitive analysis
+- `willingness_to_pay` - Payment willingness info
+
 ## Example Usage
 
 Once configured, you can ask Claude questions like:
 
+### Basic Insights
 - "Show me all negative insights from last week"
 - "What are the top customer insights this month?"
-- "Find insights about the dashboard being confusing"
-- "Give me insights on user onboarding issues"
+
+### Advanced Search
+- "Search for mobile app performance issues"
+- "Find insights about slow loading times in the Performance Issue category"
+- "Search for pricing concerns in pain points and willingness to pay fields"
+- "Find all high severity bugs from this month"
+- "Search for integration problems with Typeform"
+
+### Search Examples
+- **Phrase search**: "Find insights about 'mobile app startup'"
+- **Field-specific**: "Search pain points for 'frustrating experience'"
+- **Category filter**: "Find all Feature Request insights about 'real-time collaboration'"
+- **Severity filter**: "Search for critical issues with 'security'"
+- **Date range**: "Find performance issues from January 2024"
 
 ## License
 

package/index.js

@@ -18,7 +18,7 @@ if (!CLOSEDLOOP_API_KEY) {
 const server = new Server(
   {
     name: 'closedloop-mcp-server',
-    version: '1.1.0',
+    version: '1.1.2',
     description: 'Provides access to ClosedLoop AI product feedback data and insights with advanced search capabilities.',
   },
   {
@@ -178,7 +178,16 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
         if (!args.query) {
           throw new Error('query is required');
         }
-        response = await axios.post(`${CLOSEDLOOP_SERVER_URL}/search`, args, {
+        // Use MCP protocol for search_insights
+        response = await axios.post(CLOSEDLOOP_SERVER_URL, {
+          jsonrpc: '2.0',
+          id: 1,
+          method: 'tools/call',
+          params: {
+            name: 'search_insights',
+            arguments: args
+          }
+        }, {
           headers
         });
         break;
@@ -187,11 +196,17 @@ server.setRequestHandler(CallToolRequestSchema, async (request) => {
         throw new Error(`Unknown tool: ${name}`);
     }
 
+    // Handle MCP protocol response for search_insights
+    let responseData = response.data;
+    if (name === 'search_insights' && responseData.result) {
+      responseData = responseData.result;
+    }
+
     return {
       content: [
         {
           type: 'text',
-          text: JSON.stringify(response.data, null, 2)
+          text: JSON.stringify(responseData, null, 2)
         }
       ]
     };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@closedloop-ai/mcp-client",
-  "version": "1.1.0",
+  "version": "1.1.2",
   "description": "ClosedLoop MCP Server for AI client integration with advanced search capabilities",
   "main": "index.js",
   "bin": {