consult-llm-mcp 2.4.1 → 2.4.2

package/README.md

@@ -1,7 +1,7 @@
 # Consult LLM MCP
 
 An MCP server that lets Claude Code consult stronger AI models (GPT-5.2, Gemini
-3.0 Pro, DeepSeek Reasoner) when Sonnet has you running in circles and you need
+3.1 Pro, DeepSeek Reasoner) when Sonnet has you running in circles and you need
 to bring in the heavy artillery. Supports multi-turn conversations.
 
 ```
@@ -27,7 +27,7 @@ to bring in the heavy artillery. Supports multi-turn conversations.
 
 ## Features
 
-- Query powerful AI models (GPT-5.2, Gemini 3.0 Pro, DeepSeek Reasoner) with
+- Query powerful AI models (GPT-5.2, Gemini 3.1 Pro, DeepSeek Reasoner) with
   relevant files as context
 - Direct queries with optional file context
 - Include git changes for code review and analysis
@@ -333,7 +333,8 @@ context, but it helps.
 #### Gemini CLI
 
 Use Gemini's local CLI to take advantage of Google's
-[free quota](https://developers.google.com/gemini-code-assist/resources/quotas#quotas-for-agent-mode-gemini-cli).
+[free quota](https://developers.google.com/gemini-code-assist/resources/quotas#quotas-for-agent-mode-gemini-cli)
+or use your Google AI Pro subscription.
 
 **Requirements:**
 
@@ -456,7 +457,8 @@ See the "Using web mode..." example above for a concrete transcript.
 - `DEEPSEEK_API_KEY` - Your DeepSeek API key (required for DeepSeek models)
 - `CONSULT_LLM_DEFAULT_MODEL` - Override the default model (optional)
   - Options: `gpt-5.2` (default), `gemini-2.5-pro`, `gemini-3-pro-preview`,
-    `deepseek-reasoner`, `gpt-5.3-codex`, `gpt-5.2-codex`
+    `gemini-3.1-pro-preview`, `deepseek-reasoner`, `gpt-5.3-codex`,
+    `gpt-5.2-codex`
 - `GEMINI_BACKEND` - Backend for Gemini models (optional)
   - Options: `api` (default), `gemini-cli`, `cursor-cli`
 - `OPENAI_BACKEND` - Backend for OpenAI models (optional)
@@ -554,7 +556,8 @@ models complex questions.
 
 - **model** (optional): LLM model to use
   - Options: `gpt-5.2` (default), `gemini-2.5-pro`, `gemini-3-pro-preview`,
-    `deepseek-reasoner`, `gpt-5.3-codex`, `gpt-5.2-codex`
+    `gemini-3.1-pro-preview`, `deepseek-reasoner`, `gpt-5.3-codex`,
+    `gpt-5.2-codex`
 
 - **task_mode** (optional): Controls the system prompt persona. The calling LLM
   should choose based on the task:
@@ -589,6 +592,8 @@ models complex questions.
 - **gemini-2.5-pro**: Google's Gemini 2.5 Pro ($1.25/$10 per million tokens)
 - **gemini-3-pro-preview**: Google's Gemini 3 Pro Preview ($2/$12 per million
   tokens for prompts ≤200k tokens, $4/$18 for prompts >200k tokens)
+- **gemini-3.1-pro-preview**: Google's Gemini 3.1 Pro Preview ($2/$12 per
+  million tokens for prompts ≤200k tokens, $4/$18 for prompts >200k tokens)
 - **deepseek-reasoner**: DeepSeek's reasoning model ($0.55/$2.19 per million
   tokens)
 - **gpt-5.2**: OpenAI's latest GPT model

package/dist/config.d.ts

@@ -1,4 +1,12 @@
 import { z } from 'zod/v4';
+export interface ProviderAvailability {
+    geminiApiKey?: string;
+    geminiBackend: string;
+    openaiApiKey?: string;
+    openaiBackend: string;
+    deepseekApiKey?: string;
+}
+export declare function filterByAvailability(models: string[], providers: ProviderAvailability): string[];
 /** Build the final model catalog from built-in + extra + allowlist filtering. */
 export declare function buildModelCatalog(builtinModels: readonly string[], extraModelsRaw?: string, allowedModelsRaw?: string): string[];
 export declare const SupportedChatModel: z.ZodEnum<{

package/dist/config.js

@@ -1,6 +1,21 @@
 import { z } from 'zod/v4';
 import { ALL_MODELS } from './models.js';
 import { logToFile } from './logger.js';
+export function filterByAvailability(models, providers) {
+    return models.filter((model) => {
+        if (model.startsWith('gemini-')) {
+            return providers.geminiBackend !== 'api' || !!providers.geminiApiKey;
+        }
+        if (model.startsWith('gpt-')) {
+            return providers.openaiBackend !== 'api' || !!providers.openaiApiKey;
+        }
+        if (model.startsWith('deepseek-')) {
+            return !!providers.deepseekApiKey;
+        }
+        // Unknown prefix (user-added extra models) — always include
+        return true;
+    });
+}
 /** Build the final model catalog from built-in + extra + allowlist filtering. */
 export function buildModelCatalog(builtinModels, extraModelsRaw, allowedModelsRaw) {
     const extraModels = extraModelsRaw
@@ -23,9 +38,20 @@ export function buildModelCatalog(builtinModels, extraModelsRaw, allowedModelsRa
         ? allAvailable.filter((m) => allowedModels.includes(m))
         : allAvailable;
 }
-const enabledModels = buildModelCatalog(ALL_MODELS, process.env.CONSULT_LLM_EXTRA_MODELS, process.env.CONSULT_LLM_ALLOWED_MODELS);
+// Resolve backends early (needed for availability filtering)
+const resolvedGeminiBackend = migrateBackendEnv(process.env.GEMINI_BACKEND, process.env.GEMINI_MODE, 'gemini-cli', 'GEMINI_MODE', 'GEMINI_BACKEND');
+const resolvedOpenaiBackend = migrateBackendEnv(process.env.OPENAI_BACKEND, process.env.OPENAI_MODE, 'codex-cli', 'OPENAI_MODE', 'OPENAI_BACKEND');
+// Build catalog, then filter to only available providers
+const catalogModels = buildModelCatalog(ALL_MODELS, process.env.CONSULT_LLM_EXTRA_MODELS, process.env.CONSULT_LLM_ALLOWED_MODELS);
+const enabledModels = filterByAvailability(catalogModels, {
+    geminiApiKey: process.env.GEMINI_API_KEY,
+    geminiBackend: resolvedGeminiBackend ?? 'api',
+    openaiApiKey: process.env.OPENAI_API_KEY,
+    openaiBackend: resolvedOpenaiBackend ?? 'api',
+    deepseekApiKey: process.env.DEEPSEEK_API_KEY,
+});
 if (enabledModels.length === 0) {
-    const msg = 'Invalid environment variables:\n  CONSULT_LLM_ALLOWED_MODELS: No valid models enabled.';
+    const msg = 'Invalid environment variables:\n  No models available. Set API keys or configure CLI backends.';
     logToFile(`FATAL ERROR:\n${msg}`);
     console.error(`❌ ${msg}`);
     process.exit(1);
@@ -62,8 +88,8 @@ const parsedConfig = Config.safeParse({
     geminiApiKey: process.env.GEMINI_API_KEY,
     deepseekApiKey: process.env.DEEPSEEK_API_KEY,
     defaultModel: process.env.CONSULT_LLM_DEFAULT_MODEL,
-    geminiBackend: migrateBackendEnv(process.env.GEMINI_BACKEND, process.env.GEMINI_MODE, 'gemini-cli', 'GEMINI_MODE', 'GEMINI_BACKEND'),
-    openaiBackend: migrateBackendEnv(process.env.OPENAI_BACKEND, process.env.OPENAI_MODE, 'codex-cli', 'OPENAI_MODE', 'OPENAI_BACKEND'),
+    geminiBackend: resolvedGeminiBackend,
+    openaiBackend: resolvedOpenaiBackend,
     codexReasoningEffort: process.env.CODEX_REASONING_EFFORT,
     systemPromptPath: process.env.CONSULT_LLM_SYSTEM_PROMPT_PATH,
 });

package/dist/config.test.js

@@ -1,5 +1,5 @@
 import { describe, it, expect, vi } from 'vitest';
-import { migrateBackendEnv, buildModelCatalog } from './config.js';
+import { migrateBackendEnv, buildModelCatalog, filterByAvailability, } from './config.js';
 import { ALL_MODELS } from './models.js';
 vi.mock('./logger.js', () => ({ logToFile: vi.fn() }));
 describe('migrateBackendEnv', () => {
@@ -19,6 +19,94 @@ describe('migrateBackendEnv', () => {
         expect(migrateBackendEnv(undefined, 'cli', 'codex-cli', 'OPENAI_MODE', 'OPENAI_BACKEND')).toBe('codex-cli');
     });
 });
+describe('filterByAvailability', () => {
+    const allModels = ['gemini-2.5-pro', 'gpt-5.2', 'deepseek-reasoner'];
+    it('includes gemini models when using gemini-cli backend', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'gemini-cli',
+            openaiBackend: 'api',
+        });
+        expect(result).toContain('gemini-2.5-pro');
+        expect(result).not.toContain('gpt-5.2');
+    });
+    it('includes gemini models when using cursor-cli backend', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'cursor-cli',
+            openaiBackend: 'api',
+        });
+        expect(result).toContain('gemini-2.5-pro');
+    });
+    it('includes gemini models when API key is set', () => {
+        const result = filterByAvailability(allModels, {
+            geminiApiKey: 'key',
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        });
+        expect(result).toContain('gemini-2.5-pro');
+    });
+    it('excludes gemini models when backend is api and no key', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        });
+        expect(result).not.toContain('gemini-2.5-pro');
+    });
+    it('includes gpt models when using codex-cli backend', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'codex-cli',
+        });
+        expect(result).toContain('gpt-5.2');
+        expect(result).not.toContain('gemini-2.5-pro');
+    });
+    it('includes gpt models when using cursor-cli backend', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'cursor-cli',
+        });
+        expect(result).toContain('gpt-5.2');
+    });
+    it('includes gpt models when API key is set', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiApiKey: 'key',
+            openaiBackend: 'api',
+        });
+        expect(result).toContain('gpt-5.2');
+    });
+    it('excludes gpt models when backend is api and no key', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        });
+        expect(result).not.toContain('gpt-5.2');
+    });
+    it('includes deepseek models only when API key is set', () => {
+        expect(filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+            deepseekApiKey: 'key',
+        })).toContain('deepseek-reasoner');
+        expect(filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        })).not.toContain('deepseek-reasoner');
+    });
+    it('includes unknown-prefix models (user extras)', () => {
+        const result = filterByAvailability([...allModels, 'grok-3'], {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        });
+        expect(result).toContain('grok-3');
+    });
+    it('returns empty array when no providers are available', () => {
+        const result = filterByAvailability(allModels, {
+            geminiBackend: 'api',
+            openaiBackend: 'api',
+        });
+        expect(result).toEqual([]);
+    });
+});
 describe('buildModelCatalog', () => {
     it('returns all built-in models when no env vars are set', () => {
         const result = buildModelCatalog(ALL_MODELS);

package/dist/llm-cost.js

@@ -11,6 +11,10 @@ const MODEL_PRICING = {
         inputCostPerMillion: 2.0,
         outputCostPerMillion: 12.0,
     },
+    'gemini-3.1-pro-preview': {
+        inputCostPerMillion: 2.0,
+        outputCostPerMillion: 12.0,
+    },
     'deepseek-reasoner': {
         inputCostPerMillion: 0.55,
         outputCostPerMillion: 2.19,

package/dist/models.d.ts

@@ -1 +1 @@
-export declare const ALL_MODELS: readonly ["gemini-2.5-pro", "gemini-3-pro-preview", "deepseek-reasoner", "gpt-5.2", "gpt-5.3-codex", "gpt-5.2-codex"];
+export declare const ALL_MODELS: readonly ["gemini-2.5-pro", "gemini-3-pro-preview", "gemini-3.1-pro-preview", "deepseek-reasoner", "gpt-5.2", "gpt-5.3-codex", "gpt-5.2-codex"];

package/dist/models.js

@@ -1,6 +1,7 @@
 export const ALL_MODELS = [
     'gemini-2.5-pro',
     'gemini-3-pro-preview',
+    'gemini-3.1-pro-preview',
     'deepseek-reasoner',
     'gpt-5.2',
     'gpt-5.3-codex',

package/dist/version.d.ts

@@ -1 +1 @@
-export declare const GIT_HASH = "bbaa0d5";
+export declare const GIT_HASH = "1c9e0da";

package/dist/version.js

@@ -1 +1 @@
-export const GIT_HASH = "bbaa0d5";
+export const GIT_HASH = "1c9e0da";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "consult-llm-mcp",
-  "version": "2.4.1",
+  "version": "2.4.2",
   "description": "MCP server for consulting powerful AI models",
   "type": "module",
   "main": "dist/main.js",