converse-mcp-server 2.22.1 → 2.22.3

package/.env.example

@@ -77,6 +77,10 @@ OPENROUTER_REFERER=https://github.com/FallDownTheSystem/converse
 # WARNING: Interactive policies may cause hangs in server/headless mode
 # CODEX_APPROVAL_POLICY=never
 
+# Default Codex model (default: gpt-5.5)
+# Options: gpt-5.5, gpt-5.3-codex, gpt-5-codex
+# CODEX_MODEL=gpt-5.5
+
 # ============================================
 # Server Configuration
 # ============================================

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "converse-mcp-server",
-  "version": "2.22.1",
+  "version": "2.22.3",
   "description": "Converse MCP Server - Converse with other LLMs with chat and consensus tools",
   "type": "module",
   "main": "src/index.js",
@@ -94,14 +94,14 @@
     ".env.example"
   ],
   "dependencies": {
-    "@anthropic-ai/claude-agent-sdk": "^0.2.110",
-    "@anthropic-ai/sdk": "^0.86.1",
+    "@anthropic-ai/claude-agent-sdk": "^0.2.118",
+    "@anthropic-ai/sdk": "^0.90.0",
     "@github/copilot-sdk": "^0.2.2",
     "@google/genai": "^1.50.1",
-    "@mistralai/mistralai": "^2.2.0",
+    "@mistralai/mistralai": "^2.2.1",
     "@modelcontextprotocol/sdk": "^1.29.0",
-    "@openai/codex-sdk": "^0.118.0",
-    "ai": "^6.0.164",
+    "@openai/codex-sdk": "^0.123.0",
+    "ai": "^6.0.168",
     "ai-sdk-provider-gemini-cli": "^2.0.1",
     "cors": "^2.8.6",
     "dotenv": "^17.4.2",
@@ -110,14 +110,14 @@
     "nanoid": "^5.1.9",
     "openai": "^6.34.0",
     "p-limit": "^7.3.0",
-    "vite": "^8.0.8"
+    "vite": "^8.0.10"
   },
   "devDependencies": {
-    "@vitest/coverage-v8": "^4.1.4",
+    "@vitest/coverage-v8": "^4.1.5",
     "cross-env": "^10.1.0",
-    "eslint": "^10.2.0",
+    "eslint": "^10.2.1",
     "prettier": "^3.8.3",
     "rimraf": "^6.1.3",
-    "vitest": "^4.1.4"
+    "vitest": "^4.1.5"
   }
 }

package/src/config.js

@@ -276,6 +276,12 @@ const CONFIG_SCHEMA = {
       description:
         'Approval policy (never | untrusted | on-failure | on-request)',
     },
+    CODEX_MODEL: {
+      type: 'string',
+      default: 'gpt-5.5',
+      description:
+        'Default Codex model (e.g., gpt-5.5, gpt-5.3-codex, gpt-5-codex)',
+    },
 
     // Copilot configuration
     COPILOT_TOOL_ACCESS: {

package/src/providers/codex.js

@@ -21,22 +21,24 @@ import { normalizeExtendedPath } from '../utils/pathUtils.js';
 const SUPPORTED_MODELS = {
   codex: {
     modelName: 'codex',
-    friendlyName: 'OpenAI Codex (GPT-5.3)',
+    friendlyName: 'OpenAI Codex (GPT-5.5)',
     contextWindow: 400000,
     maxOutputTokens: 128000,
     supportsStreaming: true,
-    supportsImages: false, // Codex doesn't support images
+    supportsImages: true, // Codex SDK 0.118+ supports images via --image (local_image input)
     supportsTemperature: false, // Codex manages temperature internally
     supportsWebSearch: false, // Codex accesses files directly, not web
     timeout: 600000, // 10 minutes
     description:
-      'OpenAI Codex agentic coding assistant with local file access and tool execution (GPT-5.3-Codex)',
+      'OpenAI Codex agentic coding assistant with local file access and tool execution (GPT-5.5)',
     aliases: [
       'gpt-5-codex',
       'gpt5-codex',
       'gpt-5.2-codex',
       'gpt-5.3-codex',
       'gpt5.3-codex',
+      'gpt-5.5',
+      'gpt5.5',
     ],
   },
 };
@@ -91,14 +93,19 @@ async function getCodexSDK() {
 }
 
 /**
- * Convert message array to single prompt for Codex
- * Codex expects single prompts, not message history
+ * Convert message array to Codex SDK Input (string | UserInput[])
+ * Codex expects single prompts (new thread) or incremental input (resumed thread);
+ * history is managed SDK-side.
  *
- * Strategy:
- * - For new threads: Extract last user message only
- * - For resumed threads: Same - Codex maintains history internally
+ * Returns a plain string when the last user message is text-only, or an array
+ * of { type: 'text' | 'local_image' } parts when images are present. The SDK
+ * passes local_image paths to the CLI via --image.
+ *
+ * Images must be on-disk files — Converse stores the original path in
+ * metadata.path (chat.js / consensus.js set includeMetadata: true). Images
+ * without a path (e.g. pasted base64 with no metadata) are skipped.
  */
-function convertMessagesToPrompt(messages) {
+function convertMessagesToCodexInput(messages) {
   if (!Array.isArray(messages)) {
     throw new CodexProviderError(
       'Messages must be an array',
@@ -113,7 +120,6 @@ function convertMessagesToPrompt(messages) {
     );
   }
 
-  // Find last user message
   const lastUserMessage = messages.filter((m) => m.role === 'user').pop();
 
   if (!lastUserMessage) {
@@ -123,28 +129,46 @@ function convertMessagesToPrompt(messages) {
     );
   }
 
-  // Extract text content from message
   if (typeof lastUserMessage.content === 'string') {
     return lastUserMessage.content;
   }
 
-  // Handle array content (multimodal format)
   if (Array.isArray(lastUserMessage.content)) {
-    const textParts = lastUserMessage.content
-      .filter((item) => item.type === 'text')
-      .map((item) => item.text);
+    const parts = [];
+    let droppedImages = 0;
+    for (const item of lastUserMessage.content) {
+      if (item.type === 'text' && item.text) {
+        parts.push({ type: 'text', text: item.text });
+      } else if (item.type === 'image') {
+        const imagePath = item.metadata?.path || item.metadata?.originalPath;
+        if (imagePath) {
+          parts.push({ type: 'local_image', path: imagePath });
+        } else {
+          droppedImages += 1;
+        }
+      }
+    }
 
-    // Log warning if images present (Codex doesn't support images)
-    const hasImages = lastUserMessage.content.some(
-      (item) => item.type === 'image',
-    );
-    if (hasImages) {
+    if (droppedImages > 0) {
       debugLog(
-        '[Codex] Warning: Images in message will be ignored (Codex does not support multimodal input)',
+        `[Codex] Skipped ${droppedImages} image(s) without a file path — Codex requires on-disk images`,
       );
     }
 
-    return textParts.join('\n');
+    if (parts.length === 0) {
+      throw new CodexProviderError(
+        'Message contained no usable text or image parts',
+        ErrorCodes.INVALID_MESSAGES,
+      );
+    }
+
+    // Collapse to plain string when there are no images — keeps the non-image
+    // path identical to the legacy behavior and slightly simpler for the SDK.
+    if (parts.every((p) => p.type === 'text')) {
+      return parts.map((p) => p.text).join('\n');
+    }
+
+    return parts;
   }
 
   throw new CodexProviderError(
@@ -153,6 +177,18 @@ function convertMessagesToPrompt(messages) {
   );
 }
 
+/**
+ * Extract the combined text from a Codex SDK Input for prompt-based checks
+ * like $imagegen detection.
+ */
+function extractPromptText(input) {
+  if (typeof input === 'string') return input;
+  return input
+    .filter((p) => p.type === 'text')
+    .map((p) => p.text)
+    .join('\n\n');
+}
+
 /**
  * Get thread ID from continuation metadata
  * Codex thread IDs are stored in continuation store for resumption
@@ -185,12 +221,14 @@ function mapReasoningEffort(effort) {
 }
 
 /**
- * Create stream generator for Codex streaming responses
- * Yields raw Codex SDK events that will be normalized by ProviderStreamNormalizer
+ * Create stream generator for Codex streaming responses.
+ * `input` is the Codex SDK Input (string | UserInput[]) — strings for plain
+ * text turns, arrays when images are attached.
+ * Yields raw Codex SDK events that will be normalized by ProviderStreamNormalizer.
  */
-async function* createStreamingGenerator(thread, prompt, signal) {
+async function* createStreamingGenerator(thread, input, signal) {
   try {
-    const { events } = await thread.runStreamed(prompt, { signal });
+    const { events } = await thread.runStreamed(input, { signal });
 
     for await (const event of events) {
       // Check for cancellation
@@ -256,8 +294,9 @@ export const codexProvider = {
       // Get Codex SDK
       const Codex = await getCodexSDK();
 
-      // Convert messages to prompt
-      const prompt = convertMessagesToPrompt(messages);
+      // Convert messages to Codex SDK input (string or structured parts with images)
+      const input = convertMessagesToCodexInput(messages);
+      const promptText = extractPromptText(input);
 
       // Get thread ID if resuming conversation
       const threadId =
@@ -289,7 +328,7 @@ export const codexProvider = {
       // into image generation via $imagegen — otherwise Codex can't save the
       // generated file. Leave higher modes (workspace-write, danger-full-access)
       // alone so an explicit user choice is never downgraded or escalated.
-      const wantsImageGen = /\$imagegen\b/i.test(prompt);
+      const wantsImageGen = /\$imagegen\b/i.test(promptText);
       const sandboxMode =
         wantsImageGen && configuredSandboxMode === 'read-only'
           ? 'workspace-write'
@@ -307,6 +346,7 @@ export const codexProvider = {
 
       // Create or resume thread
       const threadOptions = {
+        model: config.providers?.codexmodel,
         workingDirectory,
         sandboxMode,
         skipGitRepoCheck,
@@ -324,12 +364,12 @@ export const codexProvider = {
       // WORKAROUND: SDK's thread.run() hangs due to missing break after turn.completed
       // Always use streaming internally, consume synchronously when stream=false
       if (stream) {
-        return createStreamingGenerator(thread, prompt, signal);
+        return createStreamingGenerator(thread, input, signal);
       }
 
       // Synchronous mode: consume streaming internally and return complete response
       const startTime = Date.now();
-      const generator = createStreamingGenerator(thread, prompt, signal);
+      const generator = createStreamingGenerator(thread, input, signal);
 
       let content = '';
       let usage = null;