converse-mcp-server 3.5.0 → 3.6.1

package/README.md

@@ -293,7 +293,7 @@ Any other model works via its full `provider/model` slug or the `openrouter:` na
 ### Claude Agent SDK Models
 
 - **claude** (aliases: `claude-sdk`, `claude-code`): Claude via the Claude Agent SDK
-  - Defaults to Claude Fable 5; pick a specific model with `claude:fable` or `claude:opus`
+  - Defaults to Claude Fable 5.1 (`claude-fable-5-1`); `claude:fable` and `claude:fable-5.1` select 5.1, `claude:fable-5` selects 5.0, and `claude:opus` selects Opus 5
   - Uses Claude Code CLI authentication (`claude login`) - no API key needed
   - Direct filesystem access from working directory
   - Unknown `claude:`-prefixed names pass through to the SDK (e.g. `claude:claude-sonnet-4-6`)
@@ -436,7 +436,7 @@ Use `"auto"` for automatic model selection, or specify exact models:
 "opus"; // -> claude-opus-4-8 (Anthropic API)
 
 // SDK providers (subscription-based, no API key)
-"claude"; // -> Claude Agent SDK (Claude Fable 5)
+"claude"; // -> Claude Agent SDK (Claude Fable 5.1)
 "claude:opus"; // -> Claude Agent SDK (Claude Opus 4.8)
 "copilot:gpt-5.6-terra"; // -> GitHub Copilot SDK
 ```
@@ -450,7 +450,7 @@ Provider priority order (subscription-based SDK providers first, then API-key pr
 
 1. Codex (`codex`)
 2. Gemini via Antigravity CLI (`gemini` → Gemini 3.8 Flash, `gemini:pro`)
-3. Claude Agent SDK (`claude` → Claude Fable 5)
+3. Claude Agent SDK (`claude` → Claude Fable 5.1)
 4. Copilot (`copilot`)
 5. OpenAI (`gpt-5.6`)
 6. Google (`gemini-pro`)
@@ -462,6 +462,8 @@ Provider priority order (subscription-based SDK providers first, then API-key pr
 
 The system will use the first 3 providers that are available (authenticated SDK or valid API key). This enables automatic multi-model consensus without manually specifying models.
 
+**Antigravity CLI permissions:** The `gemini`, `gemini:flash`, and `gemini:pro` aliases launch `agy` with `--dangerously-skip-permissions` because headless calls cannot prompt for tool approval. All tool permission requests are auto-approved, including shell commands and file writes; a read-only prompt is not an enforced security boundary. Use this provider only with trusted prompts and context. This also applies when `auto` selects it. The Google API provider is unaffected.
+
 ### Advanced Configuration
 
 #### Manual Installation Options

package/docs/API.md

@@ -468,11 +468,13 @@ Any other model works via its full `provider/model` slug (e.g. `anthropic/claude
 
 **Claude** is available through the Claude Agent SDK, using Claude Code CLI authentication instead of an API key:
 
-- **Model**: `claude` (aliases: `claude-sdk`, `claude-code`) — defaults to Claude Fable 5
-- **Model selection**: `claude:fable` (Claude Fable 5) or `claude:opus` (Claude Opus 5); unknown `claude:`-prefixed names pass through to the SDK (e.g. `claude:claude-sonnet-4-6`)
+- **Model**: `claude` (aliases: `claude-sdk`, `claude-code`) — defaults to Claude Fable 5.1 (`claude-fable-5-1`)
+- **Model selection**: `claude:fable` or `claude:fable-5.1` (Claude Fable 5.1), `claude:fable-5` (Claude Fable 5.0), or `claude:opus` (Claude Opus 5); unknown `claude:`-prefixed names pass through to the SDK (e.g. `claude:claude-sonnet-4-6`)
 - **Authentication**: `claude login` — no `ANTHROPIC_API_KEY` needed
 - **Direct file access**: reads files from the working directory
-- `reasoning_effort` and sampling parameters are managed by the SDK
+- **Reasoning effort**: `reasoning_effort` maps to the SDK's `effort` option: `low`, `medium`, `high`, `xhigh`, or `max`; `none` and `minimal` become `low`. Omitting it retains the SDK default.
+- **Turn limit**: SDK requests allow up to 100 turns (`maxTurns: 100`).
+- Sampling parameters are managed by the SDK.
 
 ### Gemini via Antigravity CLI (subscription)
 
@@ -522,7 +524,7 @@ Use `"auto"` for automatic selection, or specify exact models:
 "z-ai/glm-5.2:online"      // OpenRouter with web search opt-in
 "fable"                    // Anthropic API (-> claude-fable-5)
 "opus"                     // Anthropic API (-> claude-opus-5)
-"claude"                   // Claude Agent SDK (-> Claude Fable 5)
+"claude"                   // Claude Agent SDK (-> Claude Fable 5.1)
 "claude:opus"              // Claude Agent SDK (Claude Opus 5)
 "gemini"                   // Antigravity CLI (Gemini 3.8 Flash)
 "copilot:gpt-5.6-terra"    // GitHub Copilot SDK

package/docs/PROVIDERS.md

@@ -195,8 +195,9 @@ agy
 - **Setup Required**: Authenticate once with `claude login` (Claude Code CLI)
 - **Environment Variables**: None (uses Claude Code credentials)
 - **Supported Models**:
-  - `claude` (aliases: `claude-sdk`, `claude-code`) - Defaults to Claude Fable 5
-  - `claude:fable` - Claude Fable 5 explicitly
+  - `claude` (aliases: `claude-sdk`, `claude-code`) - Defaults to Claude Fable 5.1 (`claude-fable-5-1`)
+  - `claude:fable` or `claude:fable-5.1` - Claude Fable 5.1 explicitly
+  - `claude:fable-5` - Claude Fable 5.0 (`claude-fable-5`)
   - `claude:opus` - Claude Opus 5
   - Other `claude:`-prefixed names pass through to the SDK (e.g. `claude:claude-sonnet-4-6`)
 
@@ -204,7 +205,9 @@ agy
 - **Subscription Access**: Uses your Claude subscription instead of API credits
 - **Local File Access**: Reads files directly from the working directory
 - **Image Support**: Via the SDK's streaming input mode
-- **SDK-Managed Parameters**: `reasoning_effort` and other sampling parameters are managed internally by the SDK
+- **Reasoning Effort**: `reasoning_effort` maps to the SDK's `effort` option: `low`, `medium`, `high`, `xhigh`, or `max`; `none` and `minimal` become `low`. Omitting it retains the SDK default.
+- **Turn Limit**: SDK requests allow up to 100 turns (`maxTurns: 100`).
+- **SDK-Managed Parameters**: Sampling parameters are managed internally by the SDK.
 
 **Differences from Anthropic API Provider:**
 - **Authentication**: Claude Code login vs `ANTHROPIC_API_KEY`
@@ -341,7 +344,7 @@ The `models` array always holds plain model-name strings. Each string routes as
 "fable"                    // Anthropic (keyword match -> claude-fable-5)
 "opus"                     // Anthropic (keyword match -> claude-opus-5)
 "sonnet"                   // Anthropic (keyword match -> claude-sonnet-4-6)
-"claude"                   // Claude Agent SDK (defaults to Claude Fable 5)
+"claude"                   // Claude Agent SDK (defaults to Claude Fable 5.1)
 "claude:opus"              // Claude Agent SDK (Claude Opus 5)
 "gemini-2.5-pro"           // Google (keyword match)
 "grok-4.5"                 // X.AI (keyword match)
@@ -391,4 +394,4 @@ The `models` array always holds plain model-name strings. Each string routes as
 - The 8 curated slugs work out of the box; the default is `z-ai/glm-5.2`
 - Any other model works via its full `provider/model` slug or the `openrouter:` namespace — no extra configuration
 - The first request to a non-curated slug fetches its capabilities from OpenRouter's live catalog; a slug absent from that catalog fails before inference with `MODEL_NOT_FOUND`
-- Append `:online` to a slug to enable web search for that request (adds a real per-request cost)
+- Append `:online` to a slug to enable web search for that request (adds a real per-request cost)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "converse-mcp-server",
-  "version": "3.5.0",
+  "version": "3.6.1",
   "description": "Converse MCP Server - Converse with other LLMs with chat and consensus tools",
   "type": "module",
   "main": "src/index.js",
@@ -93,14 +93,14 @@
     ".env.example"
   ],
   "dependencies": {
-    "@anthropic-ai/claude-agent-sdk": "^0.3.261",
-    "@anthropic-ai/sdk": "^0.123.0",
-    "@github/copilot-sdk": "^1.0.11",
+    "@anthropic-ai/claude-agent-sdk": "^0.3.263",
+    "@anthropic-ai/sdk": "^0.124.0",
+    "@github/copilot-sdk": "^1.0.13",
     "@google/genai": "^2.21.0",
     "@lydell/node-pty": "1.2.0-beta.15",
     "@mistralai/mistralai": "^2.6.4",
     "@modelcontextprotocol/sdk": "^1.30.0",
-    "@openai/codex-sdk": "^0.153.3",
+    "@openai/codex-sdk": "^0.153.4",
     "cors": "^2.8.6",
     "dotenv": "^17.4.2",
     "express": "^5.2.1",

package/src/providers/claude.js

@@ -15,13 +15,15 @@
 
 import { debugLog, debugError } from '../utils/console.js';
 import { ProviderError, ErrorCodes, StopReasons } from './interface.js';
+import { clampReasoningEffort } from '../utils/reasoningEffort.js';
 
 // Default underlying model when the request is just "claude" (or "claude:fable")
-const DEFAULT_SDK_MODEL = 'claude-fable-5';
+const DEFAULT_SDK_MODEL = 'claude-fable-5-1';
+const SDK_EFFORT_TIERS = ['low', 'medium', 'high', 'xhigh', 'max'];
 
 // Supported Claude SDK models with their configurations
 const SUPPORTED_MODELS = {
-  fable: {
+  'fable-5': {
     modelName: 'claude-fable-5',
     friendlyName: 'Claude Fable 5 (via Agent SDK)',
     contextWindow: 1000000,
@@ -31,15 +33,30 @@ const SUPPORTED_MODELS = {
     supportsWebSearch: false, // SDK accesses files directly, not web
     timeout: 1800000, // 30 minutes
     description:
-      'Claude Fable 5 via Agent SDK (default) - requires claude login authentication',
+      'Claude Fable 5 via Agent SDK - requires claude login authentication',
+    aliases: ['claude-fable-5'],
+  },
+  fable: {
+    modelName: 'claude-fable-5-1',
+    friendlyName: 'Claude Fable 5.1 (via Agent SDK)',
+    contextWindow: 1000000,
+    maxOutputTokens: 128000,
+    supportsStreaming: true,
+    supportsImages: true,
+    supportsWebSearch: false,
+    timeout: 1800000,
+    description:
+      'Claude Fable 5.1 via Agent SDK (default) - requires claude login authentication',
     aliases: [
       'claude',
       'claude-sdk',
       'claude-code',
       'claude:fable',
-      'claude-fable-5',
       'claude-fable',
-      'fable-5',
+      'claude-fable-5-1',
+      'claude-fable-5.1',
+      'fable-5-1',
+      'fable-5.1',
     ],
   },
   opus: {
@@ -138,7 +155,7 @@ function findModelConfig(modelName) {
 
 /**
  * Resolve the requested model to the underlying SDK model ID.
- * - "claude" (and bare "claude:") defaults to Claude Fable 5
+ * - "claude" (and bare "claude:") defaults to Claude Fable 5.1
  * - "claude:fable" / "claude:opus" select the specific model
  * - Unknown names are passed through (after prefix stripping) so users can
  *   target any model ID the Agent SDK accepts (e.g. "claude:claude-sonnet-4-6")
@@ -299,19 +316,13 @@ async function* createStreamingGenerator(
   signal,
 ) {
   try {
-    // Build query options
-    // Use higher maxTurns to allow for file reading operations
     const queryOptions = {
+      ...options,
       model: options.model || DEFAULT_SDK_MODEL,
-      maxTurns: 20, // Allow multiple turns for file operations
+      maxTurns: 100,
       permissionMode: 'bypassPermissions', // Don't prompt for permissions
     };
 
-    // Add working directory if provided
-    if (options.cwd) {
-      queryOptions.cwd = options.cwd;
-    }
-
     // Pass abort controller if provided
     // Note: The SDK expects AbortController, not AbortSignal
     if (signal) {
@@ -459,13 +470,6 @@ export const claudeProvider = {
       );
     }
 
-    // Log unsupported parameters at debug level
-    if (reasoning_effort !== undefined) {
-      debugLog(
-        '[Claude SDK] Parameter "reasoning_effort" not supported by Claude SDK (ignored)',
-      );
-    }
-
     try {
       // Get Claude SDK
       const query = await getClaudeSDK();
@@ -481,7 +485,6 @@ export const claudeProvider = {
         debugLog('[Claude SDK] Using streaming input mode for image support');
       }
 
-      // Resolve requested model (claude/claude:fable -> claude-fable-5, claude:opus -> claude-opus-5)
       const sdkModel = resolveSdkModel(model);
       debugLog(`[Claude SDK] Resolved model "${model}" -> "${sdkModel}"`);
 
@@ -490,6 +493,9 @@ export const claudeProvider = {
         cwd: config.server?.client_cwd || process.cwd(),
         model: sdkModel,
       };
+      if (reasoning_effort !== undefined) {
+        sdkOptions.effort = clampReasoningEffort(reasoning_effort, SDK_EFFORT_TIERS);
+      }
 
       // Streaming mode
       if (stream) {

package/src/providers/gemini-cli.js

@@ -467,6 +467,7 @@ export async function runAgy({
   const args = [
     '-p',
     promptArg,
+    '--dangerously-skip-permissions',
     '--model',
     model,
     '--print-timeout',