converse-mcp-server 1.8.3 → 1.9.0

package/README.md

@@ -113,7 +113,7 @@ Get multiple AI models to analyze the same question simultaneously. Each model c
 {
   "prompt": "Should we use microservices or monolith architecture for our e-commerce platform?",
   "models": [
-    {"model": "o3"},
+    {"model": "gpt-5"},
     {"model": "gemini-2.5-flash"},
     {"model": "grok-4-0709"}
   ],
@@ -126,6 +126,9 @@ Get multiple AI models to analyze the same question simultaneously. Each model c
 ## 📊 Supported Models
 
 ### OpenAI Models
+- **gpt-5**: Latest flagship model (400K context, 128K output) - Superior reasoning, code generation, and analysis
+- **gpt-5-mini**: Faster, cost-efficient GPT-5 (400K context, 128K output) - Well-defined tasks, precise prompts
+- **gpt-5-nano**: Fastest, most cost-efficient GPT-5 (400K context, 128K output) - Summarization, classification
 - **o3**: Strong reasoning (200K context)
 - **o3-mini**: Fast O3 variant (200K context)  
 - **o3-pro**: Professional-grade reasoning (200K context) - EXTREMELY EXPENSIVE

package/bin/converse.js

@@ -1,32 +1,32 @@
 #!/usr/bin/env node
-
-/**
- * Converse MCP Server - CLI Entry Point
- * 
- * This script allows the MCP server to be run via npx/pnpm dlx for easy installation and execution.
- */
-
-import { fileURLToPath, pathToFileURL } from 'url';
-import { dirname, join } from 'path';
-import { createRequire } from 'module';
-
-// Get the directory of this script
-const __filename = fileURLToPath(import.meta.url);
-const __dirname = dirname(__filename);
-
-// Get the project root (parent of bin directory)
-const projectRoot = dirname(__dirname);
-
-// Import and start the server
-try {
-  const indexPath = join(projectRoot, 'src/index.js');
-  const { main } = await import(pathToFileURL(indexPath).href);
-  
-  // The main function will handle all logging appropriately based on transport type
-  await main();
-} catch (error) {
-  // For stdio transport, we must not output anything to stdout
-  // For http transport, this will be logged by the error handler in main
-  // Just exit with error code
-  process.exit(1);
+
+/**
+ * Converse MCP Server - CLI Entry Point
+ * 
+ * This script allows the MCP server to be run via npx/pnpm dlx for easy installation and execution.
+ */
+
+import { fileURLToPath, pathToFileURL } from 'url';
+import { dirname, join } from 'path';
+import { createRequire } from 'module';
+
+// Get the directory of this script
+const __filename = fileURLToPath(import.meta.url);
+const __dirname = dirname(__filename);
+
+// Get the project root (parent of bin directory)
+const projectRoot = dirname(__dirname);
+
+// Import and start the server
+try {
+  const indexPath = join(projectRoot, 'src/index.js');
+  const { main } = await import(pathToFileURL(indexPath).href);
+  
+  // The main function will handle all logging appropriately based on transport type
+  await main();
+} catch (error) {
+  // For stdio transport, we must not output anything to stdout
+  // For http transport, this will be logged by the error handler in main
+  // Just exit with error code
+  process.exit(1);
 }

package/docs/API.md

@@ -55,7 +55,7 @@ MCP_TRANSPORT=stdio npm start
     },
     "model": {
       "type": "string", 
-      "description": "AI model to use. Examples: 'auto' (recommended), 'gemini-2.5-flash', 'o3', 'grok-4-0709'. Default: 'auto'"
+      "description": "AI model to use. Examples: 'auto' (recommended), 'gemini-2.5-flash', 'gpt-5', 'grok-4-0709'. Default: 'auto'"
     },
     "files": {
       "type": "array",
@@ -82,7 +82,13 @@ MCP_TRANSPORT=stdio npm start
       "type": "string",
       "enum": ["minimal", "low", "medium", "high", "max"],
       "default": "medium", 
-      "description": "Reasoning depth for thinking models. Examples: 'minimal' (quick), 'medium' (balanced), 'high' (complex analysis)"
+      "description": "Reasoning depth for thinking models. Examples: 'minimal' (fastest, few reasoning tokens), 'low' (light analysis), 'medium' (balanced), 'high' (complex analysis)"
+    },
+    "verbosity": {
+      "type": "string",
+      "enum": ["low", "medium", "high"],
+      "default": "medium",
+      "description": "Output verbosity for GPT-5 models. Examples: 'low' (concise answers), 'medium' (balanced), 'high' (thorough explanations)"
     },
     "use_websearch": {
       "type": "boolean",

package/docs/EXAMPLES.md

@@ -48,6 +48,51 @@
 }
 ```
 
+## 🚀 GPT-5 Advanced Features
+
+### Using Minimal Reasoning for Fast Responses
+
+```json
+{
+  "tool": "chat",
+  "arguments": {
+    "prompt": "Write a simple SQL query to get all users created today",
+    "model": "gpt-5",
+    "reasoning_effort": "minimal",
+    "verbosity": "low"
+  }
+}
+```
+
+### High Verbosity for Detailed Explanations
+
+```json
+{
+  "tool": "chat",
+  "arguments": {
+    "prompt": "Explain the architecture of this authentication system and suggest improvements",
+    "model": "gpt-5",
+    "files": ["/c/Users/username/project/src/auth.js"],
+    "reasoning_effort": "high",
+    "verbosity": "high"
+  }
+}
+```
+
+### Cost-Efficient with GPT-5-nano
+
+```json
+{
+  "tool": "chat",
+  "arguments": {
+    "prompt": "Summarize the main points from this document",
+    "model": "gpt-5-nano",
+    "files": ["/c/Users/username/docs/report.md"],
+    "verbosity": "low"
+  }
+}
+```
+
 ## 🔧 Code Analysis Examples
 
 ### Single File Analysis
@@ -181,7 +226,7 @@
   "arguments": {
     "prompt": "Given our current system architecture, what's the best approach for implementing real-time notifications?",
     "models": [
-      {"model": "o3"},        // Most intelligent: Complex reasoning
+      {"model": "gpt-5"},        // Most intelligent: Superior reasoning
       {"model": "grok-4"},    // Most intelligent: Advanced analysis
       {"model": "gemini-2.5-pro"}  // Most intelligent: Deep thinking
     ],

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "converse-mcp-server",
-  "version": "1.8.3",
+  "version": "1.9.0",
   "description": "Converse MCP Server - Converse with other LLMs with chat and consensus tools",
   "type": "module",
   "main": "src/index.js",
@@ -63,7 +63,8 @@
     "validate": "node scripts/validate.js",
     "validate:fix": "node scripts/validate.js --fix",
     "validate:fast": "node scripts/validate.js --skip-tests --skip-lint",
-    "precommit": "npm run validate"
+    "precommit": "npm run validate",
+    "prepublishOnly": "node scripts/fix-line-endings.js"
   },
   "keywords": [
     "mcp",

package/src/prompts/helpPrompt.js

@@ -134,21 +134,22 @@ ${formatProviderModels('OpenRouter', allModels.openrouter)}
 ## Model Selection Tips
 
 ### For Complex Reasoning Tasks
-- **Most Intelligent**: o3, o3-pro, gemini-2.5-pro, grok-4
-- **Fast & Smart**: o3-mini, o4-mini, gemini-2.5-flash
-- **Budget-Friendly**: gpt-4o-mini, gemini-2.0-flash-lite
+- **Most Intelligent**: gpt-5, o3-pro, gemini-2.5-pro, grok-4
+- **Fast & Smart**: gpt-5-mini, o3-mini, o4-mini, gemini-2.5-flash
+- **Budget-Friendly**: gpt-5-nano, gpt-4o-mini, gemini-2.0-flash-lite
 
 ### For Quick Responses
-- **Ultra-Fast**: gemini-2.5-flash, gemini-2.0-flash, gpt-4o-mini
-- **Good Balance**: o4-mini, grok-3-fast
+- **Ultra-Fast**: gpt-5-nano, gemini-2.5-flash, gemini-2.0-flash, gpt-4o-mini
+- **Good Balance**: gpt-5-mini, o4-mini, grok-3-fast
 
 ### For Large Context Windows
 - **1M+ Tokens**: gpt-4.1 (1M), all Gemini models (1M)
+- **400K Tokens**: gpt-5 family (gpt-5, gpt-5-mini, gpt-5-nano)
 - **256K Tokens**: grok-4 series
 - **200K Tokens**: o3 series, o4-mini
 
 ### Special Features
-- **Web Search**: o3 series, o4-mini, gpt-4 series, gemini models with grounding, grok-4
+- **Web Search**: gpt-5, gpt-5-mini, o3 series, o4-mini, gpt-4 series, gemini models with grounding, grok-4
 - **Thinking Mode**: o3 series (reasoning_effort), gemini models (thinking budget)
 - **Image Support**: All models except gemini-2.0-flash-lite and grok-3 series
 

package/src/providers/openai.js

@@ -10,6 +10,48 @@ import { debugLog, debugError } from '../utils/console.js';
 
 // Define supported models with their capabilities
 const SUPPORTED_MODELS = {
+  'gpt-5': {
+    modelName: 'gpt-5',
+    friendlyName: 'OpenAI (GPT-5)',
+    contextWindow: 400000,
+    maxOutputTokens: 128000,
+    supportsStreaming: true,
+    supportsImages: true,
+    supportsTemperature: false,  // GPT-5 doesn't support temperature
+    supportsWebSearch: true,
+    supportsResponsesAPI: true,
+    timeout: 300000, // 5 minutes
+    description: 'Latest flagship model (400K context, 128K output) - Superior reasoning, code generation, and analysis',
+    aliases: ['gpt5', 'gpt 5', 'gpt-5-2025-08-07']
+  },
+  'gpt-5-mini': {
+    modelName: 'gpt-5-mini',
+    friendlyName: 'OpenAI (GPT-5-mini)',
+    contextWindow: 400000,
+    maxOutputTokens: 128000,
+    supportsStreaming: true,
+    supportsImages: true,
+    supportsTemperature: false,  // GPT-5 models don't support temperature
+    supportsWebSearch: true,
+    supportsResponsesAPI: true,
+    timeout: 180000, // 3 minutes
+    description: 'Faster, cost-efficient GPT-5 (400K context, 128K output) - Well-defined tasks, precise prompts',
+    aliases: ['gpt5-mini', 'gpt-5mini', 'gpt 5 mini', 'gpt-5-mini-2025-08-07']
+  },
+  'gpt-5-nano': {
+    modelName: 'gpt-5-nano',
+    friendlyName: 'OpenAI (GPT-5-nano)',
+    contextWindow: 400000,
+    maxOutputTokens: 128000,
+    supportsStreaming: true,
+    supportsImages: true,
+    supportsTemperature: false,  // GPT-5 models don't support temperature
+    supportsWebSearch: false,    // GPT-5-nano doesn't support web search
+    supportsResponsesAPI: true,
+    timeout: 120000, // 2 minutes
+    description: 'Fastest, most cost-efficient GPT-5 (400K context, 128K output) - Summarization, classification',
+    aliases: ['gpt5-nano', 'gpt-5nano', 'gpt 5 nano', 'gpt-5-nano-2025-08-07']
+  },
   'o3': {
     modelName: 'o3',
     friendlyName: 'OpenAI (O3)',
@@ -290,6 +332,7 @@ export const openaiProvider = {
       maxTokens = null,
       stream = false,
       reasoning_effort = 'medium',
+      verbosity = 'medium',
       use_websearch = false,
       config,
       ...otherOptions
@@ -343,13 +386,18 @@ export const openaiProvider = {
         requestPayload.temperature = Math.max(0, Math.min(2, temperature));
       }
 
-      // Add reasoning effort for thinking models (o3 series only)
-      if (resolvedModel.startsWith('o3') && reasoning_effort) {
+      // Add reasoning effort for thinking models (o3 series and GPT-5 family)
+      if ((resolvedModel.startsWith('o3') || resolvedModel.startsWith('gpt-5')) && reasoning_effort) {
         requestPayload.reasoning = { effort: reasoning_effort };
       }
+      
+      // Add verbosity for GPT-5 models
+      if (resolvedModel.startsWith('gpt-5') && verbosity) {
+        requestPayload.text = { verbosity: verbosity };
+      }
     } else {
       // Build Chat Completions API payload
-      const { reasoning_effort: _unused, ...cleanOptions } = otherOptions;
+      const { reasoning_effort: _unused, verbosity: _unused2, ...cleanOptions } = otherOptions;
       requestPayload = {
         model: resolvedModel,
         messages: openaiMessages,
@@ -362,10 +410,15 @@ export const openaiProvider = {
         requestPayload.temperature = Math.max(0, Math.min(2, temperature));
       }
 
-      // Add reasoning effort for thinking models (o3 series only)
-      if (resolvedModel.startsWith('o3') && reasoning_effort) {
+      // Add reasoning effort for thinking models (o3 series and GPT-5 family)
+      if ((resolvedModel.startsWith('o3') || resolvedModel.startsWith('gpt-5')) && reasoning_effort) {
         requestPayload.reasoning_effort = reasoning_effort;
       }
+      
+      // Add verbosity for GPT-5 models
+      if (resolvedModel.startsWith('gpt-5') && verbosity) {
+        requestPayload.verbosity = verbosity;
+      }
     }
 
     // Add max tokens if specified (both APIs)

package/src/tools/chat.js

@@ -40,7 +40,8 @@ export async function chatTool(args, dependencies) {
       temperature = 0.5,
       use_websearch = false,
       images = [],
-      reasoning_effort = 'medium'
+      reasoning_effort = 'medium',
+      verbosity = 'medium'
     } = args;
 
     let conversationHistory = [];
@@ -177,6 +178,7 @@ export async function chatTool(args, dependencies) {
       model: resolvedModel,
       temperature,
       reasoning_effort,
+      verbosity,
       use_websearch,
       config
     };
@@ -369,7 +371,7 @@ chatTool.inputSchema = {
     },
     model: {
       type: 'string',
-      description: 'AI model to use. Examples: "auto" (recommended), "o3", "gemini-2.5-pro", "grok-4-0709". Defaults to auto-selection.',
+      description: 'AI model to use. Examples: "auto" (recommended), "gpt-5", "gemini-2.5-pro", "grok-4-0709". Defaults to auto-selection.',
     },
     files: {
       type: 'array',
@@ -395,7 +397,13 @@ chatTool.inputSchema = {
     reasoning_effort: {
       type: 'string',
       enum: ['minimal', 'low', 'medium', 'high', 'max'],
-      description: 'Reasoning depth for thinking models. Examples: "low" (light analysis), "medium" (balanced), "high" (complex analysis). Default: "medium"',
+      description: 'Reasoning depth for thinking models. Examples: "minimal" (fastest, few reasoning tokens), "low" (light analysis), "medium" (balanced), "high" (complex analysis). Default: "medium"',
+      default: 'medium'
+    },
+    verbosity: {
+      type: 'string',
+      enum: ['low', 'medium', 'high'],
+      description: 'Output verbosity for GPT-5 models. Examples: "low" (concise answers), "medium" (balanced), "high" (thorough explanations). Default: "medium"',
       default: 'medium'
     },
     use_websearch: {