@pellux/goodvibes-sdk 0.18.14 → 0.18.16

package/dist/_internal/platform/providers/builtin-registry.d.ts

@@ -0,0 +1,27 @@
+import type { LLMProvider } from './interface.js';
+import { type BuiltinCompatDefinition } from './builtin-catalog.js';
+import type { BenchmarkEntry } from '@pellux/goodvibes-sdk/platform/providers/model-benchmarks';
+import type { RuntimeEventBus } from '../runtime/events/index.js';
+import type { CanonicalModel } from './synthetic.js';
+import type { SubscriptionManager } from '@pellux/goodvibes-sdk/platform/config/subscriptions';
+export interface ProviderRegistrar {
+    register(provider: LLMProvider): void;
+}
+export declare const CATALOG_PROVIDER_NAME_ALIASES: Record<string, string>;
+export declare function createBuiltinCompatProvider(definition: BuiltinCompatDefinition, apiKey: string, options: {
+    readonly cacheHitTracker?: import('@pellux/goodvibes-sdk/platform/providers/cache-strategy').CacheHitTracker;
+}): LLMProvider;
+/**
+ * Register the complete set of builtin provider implementations.
+ * Kept in a separate module so the registry composition root stays focused.
+ */
+export declare function registerBuiltinProviders(registry: ProviderRegistrar, hasProvider: (name: string) => boolean, apiKey: (name: string) => string, options: {
+    readonly resolveProvider: (providerName: string) => LLMProvider;
+    readonly cacheHitTracker?: import('@pellux/goodvibes-sdk/platform/providers/cache-strategy').CacheHitTracker;
+    readonly getCatalogModels: () => readonly CanonicalModel[];
+    readonly getBenchmarks: (modelId: string) => BenchmarkEntry | undefined;
+    readonly githubCopilotTokenCachePath: string;
+    readonly subscriptionManager: Pick<SubscriptionManager, 'get' | 'getPending' | 'saveSubscription' | 'resolveAccessToken'>;
+    readonly runtimeBus?: RuntimeEventBus | null;
+}): void;
+//# sourceMappingURL=builtin-registry.d.ts.map

package/dist/_internal/platform/providers/builtin-registry.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"builtin-registry.d.ts","sourceRoot":"","sources":["../../../../src/_internal/platform/providers/builtin-registry.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,gBAAgB,CAAC;AAWlD,OAAO,EAA4B,KAAK,uBAAuB,EAAE,MAAM,sBAAsB,CAAC;AAG9F,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,2DAA2D,CAAC;AAChG,OAAO,KAAK,EAAE,eAAe,EAAE,MAAM,4BAA4B,CAAC;AAClE,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,gBAAgB,CAAC;AACrD,OAAO,KAAK,EAAE,mBAAmB,EAAE,MAAM,qDAAqD,CAAC;AAE/F,MAAM,WAAW,iBAAiB;IAChC,QAAQ,CAAC,QAAQ,EAAE,WAAW,GAAG,IAAI,CAAC;CACvC;AAED,eAAO,MAAM,6BAA6B,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAWhE,CAAC;AAmBF,wBAAgB,2BAA2B,CACzC,UAAU,EAAE,uBAAuB,EACnC,MAAM,EAAE,MAAM,EACd,OAAO,EAAE;IACP,QAAQ,CAAC,eAAe,CAAC,EAAE,OAAO,yDAAyD,EAAE,eAAe,CAAC;CAC9G,GACA,WAAW,CA6Cb;AAED;;;GAGG;AACH,wBAAgB,wBAAwB,CACtC,QAAQ,EAAE,iBAAiB,EAC3B,WAAW,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,EACtC,MAAM,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,MAAM,EAChC,OAAO,EAAE;IACP,QAAQ,CAAC,eAAe,EAAE,CAAC,YAAY,EAAE,MAAM,KAAK,WAAW,CAAC;IAChE,QAAQ,CAAC,eAAe,CAAC,EAAE,OAAO,yDAAyD,EAAE,eAAe,CAAC;IAC7G,QAAQ,CAAC,gBAAgB,EAAE,MAAM,SAAS,cAAc,EAAE,CAAC;IAC3D,QAAQ,CAAC,aAAa,EAAE,CAAC,OAAO,EAAE,MAAM,KAAK,cAAc,GAAG,SAAS,CAAC;IACxE,QAAQ,CAAC,2BAA2B,EAAE,MAAM,CAAC;IAC7C,QAAQ,CAAC,mBAAmB,EAAE,IAAI,CAAC,mBAAmB,EAAE,KAAK,GAAG,YAAY,GAAG,kBAAkB,GAAG,oBAAoB,CAAC,CAAC;IAC1H,QAAQ,CAAC,UAAU,CAAC,EAAE,eAAe,GAAG,IAAI,CAAC;CAC9C,GACA,IAAI,CAmcN"}

package/dist/_internal/platform/providers/builtin-registry.js

@@ -0,0 +1,510 @@
+import { OpenAIProvider } from './openai.js';
+import { OpenAICompatProvider } from './openai-compat.js';
+import { AnthropicProvider } from './anthropic.js';
+import { AnthropicCompatProvider } from './anthropic-compat.js';
+import { OpenAICodexProvider } from './openai-codex.js';
+import { GeminiProvider } from './gemini.js';
+import { AmazonBedrockProvider } from './amazon-bedrock.js';
+import { AmazonBedrockMantleProvider } from './amazon-bedrock-mantle.js';
+import { AnthropicVertexProvider } from './anthropic-vertex.js';
+import { GitHubCopilotProvider } from './github-copilot.js';
+import { BUILTIN_COMPAT_PROVIDERS } from './builtin-catalog.js';
+import { normalizeFoundryEndpoint } from '@pellux/goodvibes-sdk/platform/providers/microsoft-foundry-shared';
+import { SyntheticProvider } from './synthetic.js';
+export const CATALOG_PROVIDER_NAME_ALIASES = {
+    inception: 'inceptionlabs',
+    copilot: 'github-copilot',
+    'azure-openai': 'microsoft-foundry',
+    'azure-openai-responses': 'microsoft-foundry',
+    dashscope: 'qwen',
+    'volcano-engine': 'volcengine',
+    'x-ai': 'xai',
+    'z-ai': 'zai',
+    'cloudflare-gateway': 'cloudflare-ai-gateway',
+    'ai-gateway': 'vercel-ai-gateway',
+};
+function firstEnvValue(envVars) {
+    for (const envVar of envVars) {
+        const value = process.env[envVar];
+        if (typeof value === 'string' && value.trim().length > 0) {
+            return value.trim();
+        }
+    }
+    return undefined;
+}
+function resolveFoundryBaseUrl() {
+    const endpoint = firstEnvValue(['AZURE_OPENAI_ENDPOINT', 'AZURE_FOUNDRY_ENDPOINT']);
+    if (!endpoint)
+        return 'https://example.openai.azure.com/openai/v1';
+    const normalized = normalizeFoundryEndpoint(endpoint);
+    return normalized.endsWith('/openai/v1') ? normalized : `${normalized}/openai/v1`;
+}
+export function createBuiltinCompatProvider(definition, apiKey, options) {
+    if (definition.kind === 'anthropic-compat') {
+        return new AnthropicCompatProvider({
+            name: definition.id,
+            baseURL: definition.baseURL,
+            apiKey,
+            defaultModel: definition.defaultModel,
+            models: [...definition.models],
+            ...(definition.defaultHeaders ? { defaultHeaders: definition.defaultHeaders } : {}),
+            authEnvVars: definition.envVars,
+            serviceNames: definition.serviceNames,
+            aliases: definition.aliases,
+            ...(definition.subscriptionProviderId ? { subscriptionProviderId: definition.subscriptionProviderId } : {}),
+            ...(definition.streamProtocol ? { streamProtocol: definition.streamProtocol } : {}),
+            ...(definition.authHeaderMode ? { authHeaderMode: definition.authHeaderMode } : {}),
+            ...(definition.allowAnonymous ? { allowAnonymous: true } : {}),
+            ...(definition.anonymousConfigured ? { anonymousConfigured: true } : {}),
+            ...(definition.anonymousDetail ? { anonymousDetail: definition.anonymousDetail } : {}),
+            ...(options.cacheHitTracker ? { cacheHitTracker: options.cacheHitTracker } : {}),
+        });
+    }
+    const baseURL = definition.id === 'microsoft-foundry' ? resolveFoundryBaseUrl() : definition.baseURL;
+    const effectiveApiKey = apiKey || (definition.allowAnonymous ? 'gv-local' : '');
+    return new OpenAICompatProvider({
+        name: definition.id,
+        baseURL,
+        apiKey: effectiveApiKey,
+        authConfigured: Boolean(apiKey),
+        defaultModel: definition.defaultModel,
+        models: [...definition.models],
+        ...(definition.embeddingModel ? { embeddingModel: definition.embeddingModel } : {}),
+        ...(definition.defaultHeaders ? { defaultHeaders: definition.defaultHeaders } : {}),
+        reasoningFormat: definition.reasoningFormat ?? 'none',
+        authEnvVars: definition.envVars,
+        serviceNames: definition.serviceNames,
+        ...(definition.subscriptionProviderId ? { subscriptionProviderId: definition.subscriptionProviderId } : {}),
+        ...(definition.suppressedModels ? { suppressedModels: definition.suppressedModels } : {}),
+        ...(definition.aliases ? { aliases: definition.aliases } : {}),
+        ...(definition.streamProtocol ? { streamProtocol: definition.streamProtocol } : {}),
+        ...(definition.allowAnonymous ? { allowAnonymous: true } : {}),
+        ...(definition.anonymousConfigured ? { anonymousConfigured: true } : {}),
+        ...(definition.anonymousDetail ? { anonymousDetail: definition.anonymousDetail } : {}),
+        ...(options.cacheHitTracker ? { cacheHitTracker: options.cacheHitTracker } : {}),
+    });
+}
+/**
+ * Register the complete set of builtin provider implementations.
+ * Kept in a separate module so the registry composition root stays focused.
+ */
+export function registerBuiltinProviders(registry, hasProvider, apiKey, options) {
+    registry.register(new OpenAICompatProvider({
+        name: 'inceptionlabs',
+        baseURL: 'https://api.inceptionlabs.ai/v1',
+        apiKey: apiKey('inceptionlabs'),
+        defaultModel: 'mercury-2',
+        models: ['mercury-2', 'mercury-edit'],
+        reasoningFormat: 'mercury',
+        authEnvVars: ['INCEPTION_API_KEY'],
+        serviceNames: ['inceptionlabs'],
+        aliases: ['inception'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'openrouter',
+        baseURL: 'https://openrouter.ai/api/v1',
+        apiKey: apiKey('openrouter'),
+        defaultModel: 'openrouter/free',
+        models: [
+            'openrouter/free',
+            'arcee-ai/trinity-mini:free',
+            'minimax/minimax-m2.5:free',
+            'nvidia/nemotron-3-super-120b-a12b:free',
+            'nvidia/nemotron-3-nano-30b-a3b:free',
+            'nvidia/nemotron-nano-12b-v2-vl:free',
+            'nvidia/nemotron-nano-9b-v2:free',
+            'openai/gpt-oss-120b:free',
+            'openai/gpt-oss-20b:free',
+            'stepfun/step-3.5-flash:free',
+            'z-ai/glm-4.5-air:free',
+        ],
+        reasoningFormat: 'openrouter',
+        authEnvVars: ['OPENROUTER_API_KEY'],
+        serviceNames: ['openrouter'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'aihubmix',
+        baseURL: 'https://aihubmix.com/v1',
+        apiKey: apiKey('aihubmix'),
+        defaultModel: 'gpt-4.1-free',
+        models: [
+            'gpt-4.1-free', 'gpt-4.1-mini-free', 'gpt-4.1-nano-free', 'gpt-4o-free',
+            'gemini-2.0-flash-free', 'gemini-3-flash-preview-free', 'gemini-3.1-flash-image-preview-free',
+            'glm-4.7-flash-free',
+            'coding-glm-4.6-free', 'coding-glm-4.7-free', 'coding-glm-5-free', 'coding-glm-5-turbo-free',
+            'coding-minimax-m2-free', 'coding-minimax-m2.1-free', 'coding-minimax-m2.5-free', 'coding-minimax-m2.7-free',
+            'kimi-for-coding-free', 'mimo-v2-flash-free', 'minimax-m2.5-free', 'step-3.5-flash-free',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['AIHUBMIX_API_KEY'],
+        serviceNames: ['aihubmix'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'groq',
+        baseURL: 'https://api.groq.com/openai/v1',
+        apiKey: apiKey('groq'),
+        defaultModel: 'qwen/qwen3-32b',
+        models: [
+            'qwen/qwen3-32b',
+            'openai/gpt-oss-120b', 'openai/gpt-oss-20b',
+            'moonshotai/kimi-k2-instruct', 'moonshotai/kimi-k2-instruct-0905',
+            'llama-3.3-70b-versatile', 'llama-3.1-8b-instant',
+            'meta-llama/llama-4-scout-17b-16e-instruct',
+            'groq/compound', 'groq/compound-mini',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['GROQ_API_KEY'],
+        serviceNames: ['groq'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'cerebras',
+        baseURL: 'https://api.cerebras.ai/v1',
+        apiKey: apiKey('cerebras'),
+        defaultModel: 'qwen-3-235b-a22b-instruct-2507',
+        models: ['llama3.1-8b', 'qwen-3-235b-a22b-instruct-2507'],
+        reasoningFormat: 'none',
+        authEnvVars: ['CEREBRAS_API_KEY'],
+        serviceNames: ['cerebras'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'mistral',
+        baseURL: 'https://api.mistral.ai/v1',
+        apiKey: apiKey('mistral'),
+        defaultModel: 'mistral-large-latest',
+        embeddingModel: 'mistral-embed',
+        models: [
+            'mistral-large-latest', 'mistral-medium-latest', 'mistral-small-latest',
+            'codestral-latest', 'devstral-latest', 'devstral-medium-latest', 'devstral-small-latest',
+            'magistral-medium-latest', 'magistral-small-latest',
+            'ministral-14b-latest', 'ministral-8b-latest', 'ministral-3b-latest',
+            'pixtral-large-latest', 'open-mistral-nemo',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['MISTRAL_API_KEY'],
+        serviceNames: ['mistral'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'ollama-cloud',
+        baseURL: 'https://ollama.com/v1',
+        apiKey: apiKey('ollama-cloud'),
+        defaultModel: 'deepseek-v3.2',
+        models: [
+            'deepseek-v3.2', 'deepseek-v3.1:671b', 'cogito-2.1:671b',
+            'qwen3.5:397b', 'qwen3-coder:480b', 'qwen3-coder-next', 'qwen3-next:80b',
+            'qwen3-vl:235b', 'qwen3-vl:235b-instruct',
+            'kimi-k2:1t', 'kimi-k2-thinking', 'kimi-k2.5',
+            'mistral-large-3:675b', 'devstral-2:123b', 'devstral-small-2:24b',
+            'ministral-3:14b', 'ministral-3:8b', 'ministral-3:3b',
+            'gemini-3-flash-preview', 'gemma3:27b', 'gemma3:12b', 'gemma3:4b',
+            'glm-4.6', 'glm-4.7', 'glm-5',
+            'gpt-oss:120b', 'gpt-oss:20b',
+            'minimax-m2', 'minimax-m2.1', 'minimax-m2.5', 'minimax-m2.7',
+            'nemotron-3-super', 'nemotron-3-nano:30b',
+            'rnj-1:8b',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['OLLAMA_CLOUD_API_KEY', 'OLLAMA_API_KEY'],
+        serviceNames: ['ollama-cloud'],
+    }));
+    registry.register(new OpenAIProvider(apiKey('openai'), options.cacheHitTracker));
+    registry.register(new AnthropicProvider(apiKey('anthropic'), options.cacheHitTracker));
+    registry.register(new OpenAICodexProvider(options.subscriptionManager));
+    registry.register(new GeminiProvider(apiKey('gemini'), options.cacheHitTracker));
+    registry.register(new OpenAICompatProvider({
+        name: 'huggingface',
+        baseURL: 'https://router.huggingface.co/v1',
+        apiKey: apiKey('huggingface'),
+        defaultModel: 'deepseek-ai/DeepSeek-V3.2',
+        models: [
+            'Qwen/QwQ-32B',
+            'Qwen/Qwen2.5-72B-Instruct',
+            'Qwen/Qwen2.5-7B-Instruct',
+            'Qwen/Qwen2.5-Coder-32B-Instruct',
+            'Qwen/Qwen2.5-Coder-3B-Instruct',
+            'Qwen/Qwen2.5-Coder-7B-Instruct',
+            'Qwen/Qwen2.5-VL-72B-Instruct',
+            'Qwen/Qwen2.5-VL-7B-Instruct',
+            'Qwen/Qwen3-14B',
+            'Qwen/Qwen3-235B-A22B',
+            'Qwen/Qwen3-235B-A22B-Instruct-2507',
+            'Qwen/Qwen3-235B-A22B-Thinking-2507',
+            'Qwen/Qwen3-30B-A3B',
+            'Qwen/Qwen3-32B',
+            'Qwen/Qwen3-4B-Instruct-2507',
+            'Qwen/Qwen3-4B-Thinking-2507',
+            'Qwen/Qwen3-8B',
+            'Qwen/Qwen3-Coder-30B-A3B-Instruct',
+            'Qwen/Qwen3-Coder-480B-A35B-Instruct',
+            'Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8',
+            'Qwen/Qwen3-Coder-Next',
+            'Qwen/Qwen3-Coder-Next-FP8',
+            'Qwen/Qwen3-Next-80B-A3B-Instruct',
+            'Qwen/Qwen3-Next-80B-A3B-Thinking',
+            'Qwen/Qwen3-VL-235B-A22B-Instruct',
+            'Qwen/Qwen3-VL-235B-A22B-Thinking',
+            'Qwen/Qwen3-VL-30B-A3B-Instruct',
+            'Qwen/Qwen3-VL-30B-A3B-Thinking',
+            'Qwen/Qwen3-VL-8B-Instruct',
+            'Qwen/Qwen3.5-122B-A10B',
+            'Qwen/Qwen3.5-27B',
+            'Qwen/Qwen3.5-35B-A3B',
+            'Qwen/Qwen3.5-397B-A17B',
+            'Qwen/Qwen3.5-9B',
+            'deepseek-ai/DeepSeek-Prover-V2-671B',
+            'deepseek-ai/DeepSeek-R1',
+            'deepseek-ai/DeepSeek-R1-0528',
+            'deepseek-ai/DeepSeek-R1-Distill-Llama-70B',
+            'deepseek-ai/DeepSeek-R1-Distill-Llama-8B',
+            'deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B',
+            'deepseek-ai/DeepSeek-R1-Distill-Qwen-32B',
+            'deepseek-ai/DeepSeek-R1-Distill-Qwen-7B',
+            'deepseek-ai/DeepSeek-V3',
+            'deepseek-ai/DeepSeek-V3-0324',
+            'deepseek-ai/DeepSeek-V3.1',
+            'deepseek-ai/DeepSeek-V3.1-Terminus',
+            'deepseek-ai/DeepSeek-V3.2',
+            'deepseek-ai/DeepSeek-V3.2-Exp',
+            'zai-org/AutoGLM-Phone-9B-Multilingual',
+            'zai-org/GLM-4-32B-0414',
+            'zai-org/GLM-4.5',
+            'zai-org/GLM-4.5-Air',
+            'zai-org/GLM-4.5-Air-FP8',
+            'zai-org/GLM-4.5V',
+            'zai-org/GLM-4.5V-FP8',
+            'zai-org/GLM-4.6',
+            'zai-org/GLM-4.6-FP8',
+            'zai-org/GLM-4.6V',
+            'zai-org/GLM-4.6V-FP8',
+            'zai-org/GLM-4.6V-Flash',
+            'zai-org/GLM-4.7',
+            'zai-org/GLM-4.7-FP8',
+            'zai-org/GLM-4.7-Flash',
+            'zai-org/GLM-5',
+            'meta-llama/Llama-3.1-70B-Instruct',
+            'meta-llama/Llama-3.1-8B-Instruct',
+            'meta-llama/Llama-3.2-1B-Instruct',
+            'meta-llama/Llama-3.3-70B-Instruct',
+            'meta-llama/Llama-4-Maverick-17B-128E-Instruct',
+            'meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8',
+            'meta-llama/Llama-4-Scout-17B-16E-Instruct',
+            'meta-llama/Meta-Llama-3-70B-Instruct',
+            'meta-llama/Meta-Llama-3-8B-Instruct',
+            'CohereLabs/aya-expanse-32b',
+            'CohereLabs/aya-vision-32b',
+            'CohereLabs/c4ai-command-a-03-2025',
+            'CohereLabs/c4ai-command-r-08-2024',
+            'CohereLabs/c4ai-command-r7b-12-2024',
+            'CohereLabs/c4ai-command-r7b-arabic-02-2025',
+            'CohereLabs/command-a-reasoning-08-2025',
+            'CohereLabs/command-a-translate-08-2025',
+            'CohereLabs/command-a-vision-07-2025',
+            'CohereLabs/tiny-aya-earth',
+            'CohereLabs/tiny-aya-fire',
+            'CohereLabs/tiny-aya-global',
+            'CohereLabs/tiny-aya-water',
+            'moonshotai/Kimi-K2-Instruct',
+            'moonshotai/Kimi-K2-Instruct-0905',
+            'moonshotai/Kimi-K2-Thinking',
+            'moonshotai/Kimi-K2.5',
+            'MiniMaxAI/MiniMax-M1-80k',
+            'MiniMaxAI/MiniMax-M2',
+            'MiniMaxAI/MiniMax-M2.1',
+            'MiniMaxAI/MiniMax-M2.5',
+            'google/gemma-3-27b-it',
+            'google/gemma-3n-E4B-it',
+            'openai/gpt-oss-120b',
+            'openai/gpt-oss-20b',
+            'openai/gpt-oss-safeguard-20b',
+            'XiaomiMiMo/MiMo-V2-Flash',
+            'deepcogito/cogito-671b-v2.1',
+            'deepcogito/cogito-671b-v2.1-FP8',
+            'baidu/ERNIE-4.5-21B-A3B-PT',
+            'baidu/ERNIE-4.5-300B-A47B-Base-PT',
+            'baidu/ERNIE-4.5-VL-28B-A3B-PT',
+            'baidu/ERNIE-4.5-VL-424B-A47B-Base-PT',
+            'allenai/Olmo-3-7B-Instruct',
+            'allenai/Olmo-3.1-32B-Instruct',
+            'allenai/Olmo-3.1-32B-Think',
+            'EssentialAI/rnj-1-instruct',
+            'NousResearch/Hermes-2-Pro-Llama-3-8B',
+            'Sao10K/L3-70B-Euryale-v2.1',
+            'Sao10K/L3-8B-Lunaris-v1',
+            'Sao10K/L3-8B-Stheno-v3.2',
+            'ServiceNow-AI/Apriel-1.6-15b-Thinker',
+            'aisingapore/Gemma-SEA-LION-v4-27B-IT',
+            'aisingapore/Qwen-SEA-LION-v4-32B-IT',
+            'alpindale/WizardLM-2-8x22B',
+            'dicta-il/DictaLM-3.0-24B-Thinking',
+            'katanemo/Arch-Router-1.5B',
+            'swiss-ai/Apertus-70B-Instruct-2509',
+            'swiss-ai/Apertus-8B-Instruct-2509',
+            'tokyotech-llm/Llama-3.3-Swallow-70B-Instruct-v0.4',
+            'utter-project/EuroLLM-22B-Instruct-2512',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['HF_API_KEY', 'HUGGINGFACE_API_KEY', 'HF_TOKEN'],
+        serviceNames: ['huggingface'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'nvidia',
+        baseURL: 'https://integrate.api.nvidia.com/v1',
+        apiKey: apiKey('nvidia'),
+        defaultModel: 'deepseek-ai/deepseek-v3.2',
+        models: [
+            'deepseek-ai/deepseek-v3.2',
+            'deepseek-ai/deepseek-v3.1',
+            'deepseek-ai/deepseek-v3.1-terminus',
+            'deepseek-ai/deepseek-r1-distill-qwen-32b',
+            'deepseek-ai/deepseek-r1-distill-qwen-14b',
+            'deepseek-ai/deepseek-r1-distill-qwen-7b',
+            'deepseek-ai/deepseek-r1-distill-llama-8b',
+            'deepseek-ai/deepseek-coder-6.7b-instruct',
+            'nvidia/llama-3.1-nemotron-ultra-253b-v1',
+            'nvidia/nemotron-3-super-120b-a12b',
+            'nvidia/nemotron-4-340b-instruct',
+            'nvidia/llama-3.3-nemotron-super-49b-v1.5',
+            'nvidia/llama-3.3-nemotron-super-49b-v1',
+            'nvidia/llama-3.1-nemotron-70b-instruct',
+            'nvidia/llama-3.1-nemotron-51b-instruct',
+            'nvidia/nemotron-3-nano-30b-a3b',
+            'nvidia/nemotron-nano-3-30b-a3b',
+            'nvidia/nemotron-nano-12b-v2-vl',
+            'nvidia/nvidia-nemotron-nano-9b-v2',
+            'nvidia/llama-3.1-nemotron-nano-8b-v1',
+            'nvidia/llama-3.1-nemotron-nano-vl-8b-v1',
+            'nvidia/llama-3.1-nemotron-nano-4b-v1.1',
+            'nvidia/nemotron-mini-4b-instruct',
+            'nvidia/nemotron-4-mini-hindi-4b-instruct',
+            'nvidia/usdcode-llama-3.1-70b-instruct',
+            'nvidia/llama3-chatqa-1.5-70b',
+            'nvidia/llama3-chatqa-1.5-8b',
+            'nvidia/mistral-nemo-minitron-8b-8k-instruct',
+            'nvidia/cosmos-reason2-8b',
+            'meta/llama-3.1-405b-instruct',
+            'meta/llama-3.2-90b-vision-instruct',
+            'meta/llama-3.3-70b-instruct',
+            'meta/llama-3.1-70b-instruct',
+            'meta/llama3-70b-instruct',
+            'meta/llama2-70b',
+            'meta/codellama-70b',
+            'meta/llama-4-maverick-17b-128e-instruct',
+            'meta/llama-4-scout-17b-16e-instruct',
+            'meta/llama-3.2-11b-vision-instruct',
+            'meta/llama-3.1-8b-instruct',
+            'meta/llama3-8b-instruct',
+            'meta/llama-3.2-3b-instruct',
+            'meta/llama-3.2-1b-instruct',
+            'qwen/qwen3.5-397b-a17b',
+            'qwen/qwen3-coder-480b-a35b-instruct',
+            'qwen/qwen3.5-122b-a10b',
+            'qwen/qwen3-next-80b-a3b-instruct',
+            'qwen/qwen3-next-80b-a3b-thinking',
+            'qwen/qwq-32b',
+            'qwen/qwen2.5-coder-32b-instruct',
+            'qwen/qwen2.5-coder-7b-instruct',
+            'qwen/qwen2.5-7b-instruct',
+            'qwen/qwen2-7b-instruct',
+            'moonshotai/kimi-k2.5',
+            'moonshotai/kimi-k2-thinking',
+            'moonshotai/kimi-k2-instruct',
+            'moonshotai/kimi-k2-instruct-0905',
+            'mistralai/mistral-large-3-675b-instruct-2512',
+            'mistralai/mistral-large-2-instruct',
+            'mistralai/mistral-large',
+            'mistralai/mistral-medium-3-instruct',
+            'mistralai/mistral-small-4-119b-2603',
+            'mistralai/mistral-small-3.1-24b-instruct-2503',
+            'mistralai/mistral-small-24b-instruct',
+            'mistralai/mistral-nemotron',
+            'mistralai/magistral-small-2506',
+            'mistralai/devstral-2-123b-instruct-2512',
+            'mistralai/codestral-22b-instruct-v0.1',
+            'mistralai/mamba-codestral-7b-v0.1',
+            'mistralai/mathstral-7b-v0.1',
+            'mistralai/ministral-14b-instruct-2512',
+            'mistralai/mistral-7b-instruct-v0.3',
+            'mistralai/mistral-7b-instruct-v0.2',
+            'mistralai/mixtral-8x22b-instruct-v0.1',
+            'mistralai/mixtral-8x7b-instruct-v0.1',
+            'google/gemma-3-27b-it',
+            'google/gemma-3-12b-it',
+            'google/gemma-3-4b-it',
+            'google/gemma-3-1b-it',
+            'google/gemma-3n-e4b-it',
+            'google/gemma-3n-e2b-it',
+            'google/gemma-2-27b-it',
+            'google/gemma-2-9b-it',
+            'google/gemma-2-2b-it',
+            'google/codegemma-1.1-7b',
+            'google/codegemma-7b',
+            'microsoft/phi-4-multimodal-instruct',
+            'microsoft/phi-4-mini-instruct',
+            'microsoft/phi-4-mini-flash-reasoning',
+            'microsoft/phi-3.5-moe-instruct',
+            'microsoft/phi-3.5-vision-instruct',
+            'microsoft/phi-3.5-mini-instruct',
+            'microsoft/phi-3-medium-128k-instruct',
+            'microsoft/phi-3-medium-4k-instruct',
+            'microsoft/phi-3-small-128k-instruct',
+            'microsoft/phi-3-small-8k-instruct',
+            'microsoft/phi-3-mini-128k-instruct',
+            'microsoft/phi-3-mini-4k-instruct',
+            'microsoft/phi-3-vision-128k-instruct',
+            'openai/gpt-oss-120b',
+            'openai/gpt-oss-20b',
+            'ibm/granite-34b-code-instruct',
+            'ibm/granite-3.3-8b-instruct',
+            'ibm/granite-3.0-8b-instruct',
+            'ibm/granite-3.0-3b-a800m-instruct',
+            'ibm/granite-8b-code-instruct',
+            'z-ai/glm5',
+            'z-ai/glm4.7',
+            'minimaxai/minimax-m2.5',
+            'bytedance/seed-oss-36b-instruct',
+            'stepfun-ai/step-3.5-flash',
+            'writer/palmyra-creative-122b',
+            'writer/palmyra-fin-70b-32k',
+            'writer/palmyra-med-70b',
+            'writer/palmyra-med-70b-32k',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['NVIDIA_API_KEY'],
+        serviceNames: ['nvidia'],
+    }));
+    registry.register(new OpenAICompatProvider({
+        name: 'llm7',
+        baseURL: 'https://api.llm7.io/v1',
+        apiKey: apiKey('llm7'),
+        defaultModel: 'codestral-latest',
+        models: [
+            'GLM-4.6V-Flash',
+            'codestral-latest',
+            'gpt-oss-20b',
+            'meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo',
+            'ministral-8b-2512',
+        ],
+        reasoningFormat: 'none',
+        authEnvVars: ['LLM7_API_KEY'],
+        serviceNames: ['llm7'],
+    }));
+    for (const definition of BUILTIN_COMPAT_PROVIDERS) {
+        if (hasProvider(definition.id))
+            continue;
+        const resolvedKey = apiKey(definition.id);
+        registry.register(createBuiltinCompatProvider(definition, resolvedKey, options));
+    }
+    registry.register(new AmazonBedrockProvider());
+    registry.register(new AmazonBedrockMantleProvider());
+    registry.register(new AnthropicVertexProvider());
+    registry.register(new GitHubCopilotProvider({
+        tokenCachePath: options.githubCopilotTokenCachePath,
+    }));
+    // Synthetic failover provider — must be after all backends.
+    // Stage 3: catalog-driven SyntheticProvider manages its own backend lists.
+    registry.register(new SyntheticProvider({
+        resolveProvider: options.resolveProvider,
+        getCatalogModels: options.getCatalogModels,
+        getBenchmarks: options.getBenchmarks,
+        runtimeBus: options.runtimeBus,
+    }));
+}

package/dist/_internal/platform/providers/cache-planner.d.ts

@@ -0,0 +1,58 @@
+/**
+ * Cache Strategy Planner — uses the helper model (or main model) to optimize
+ * cache breakpoint placement for explicit-caching providers.
+ *
+ * Two modes:
+ *   1. Heuristic (default): Uses getDefaultStrategy() from cache-strategy.ts
+ *   2. LLM-assisted: Sends context to the helper model for optimized strategy
+ *
+ * The planner runs:
+ *   - Once at session start (first turn)
+ *   - Every N turns (configurable, default 10)
+ *   - When cache hit rate drops below threshold
+ */
+import type { CacheStrategy, CacheContext, CacheHitTracker } from '@pellux/goodvibes-sdk/platform/providers/cache-strategy';
+import type { ConfigManager } from '../config/manager.js';
+import type { HelperModel } from '../config/helper-model.js';
+/** Result of a strategy planning run. */
+export interface PlanResult {
+    strategy: CacheStrategy;
+    source: 'heuristic' | 'helper' | 'cached';
+    planTimeMs: number;
+}
+/**
+ * CachePlanner — manages cache strategy lifecycle.
+ *
+ * Caches the current strategy and refreshes it based on
+ * turn count and hit rate thresholds.
+ */
+export declare class CachePlanner {
+    private readonly configManager;
+    private readonly helperModel;
+    private currentStrategy;
+    private lastPlanTurn;
+    private turnsSinceLastPlan;
+    private readonly cacheHitTracker;
+    constructor(configManager: Pick<ConfigManager, 'get'>, helperModel: Pick<HelperModel, 'chat'>, cacheHitTracker: Pick<CacheHitTracker, 'getMetrics'>);
+    /**
+     * Get the current cache strategy, planning a new one if needed.
+     *
+     * Triggers re-planning when:
+     *   - No strategy exists yet (first call)
+     *   - refreshAfterTurns threshold reached
+     *   - Cache hit rate dropped below warning threshold
+     */
+    getStrategy(context: CacheContext): Promise<PlanResult>;
+    /** Check if strategy needs refresh. */
+    private shouldRefresh;
+    /**
+     * Use the helper model to plan an optimized cache strategy.
+     * Returns null if the helper can't produce a valid strategy.
+     */
+    private planWithHelper;
+    /** Build the prompt for the helper model. */
+    private buildHelperPrompt;
+    /** Force a strategy refresh on the next getStrategy() call. */
+    invalidate(): void;
+}
+//# sourceMappingURL=cache-planner.d.ts.map

package/dist/_internal/platform/providers/cache-planner.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"cache-planner.d.ts","sourceRoot":"","sources":["../../../../src/_internal/platform/providers/cache-planner.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;GAYG;AAEH,OAAO,KAAK,EACV,aAAa,EACb,YAAY,EAEZ,eAAe,EAChB,MAAM,yDAAyD,CAAC;AAIjE,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAC1D,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,2BAA2B,CAAC;AAI7D,yCAAyC;AACzC,MAAM,WAAW,UAAU;IACzB,QAAQ,EAAE,aAAa,CAAC;IACxB,MAAM,EAAE,WAAW,GAAG,QAAQ,GAAG,QAAQ,CAAC;IAC1C,UAAU,EAAE,MAAM,CAAC;CACpB;AAED;;;;;GAKG;AACH,qBAAa,YAAY;IAOrB,OAAO,CAAC,QAAQ,CAAC,aAAa;IAC9B,OAAO,CAAC,QAAQ,CAAC,WAAW;IAP9B,OAAO,CAAC,eAAe,CAA8B;IACrD,OAAO,CAAC,YAAY,CAAK;IACzB,OAAO,CAAC,kBAAkB,CAAK;IAC/B,OAAO,CAAC,QAAQ,CAAC,eAAe,CAAsC;gBAGnD,aAAa,EAAE,IAAI,CAAC,aAAa,EAAE,KAAK,CAAC,EACzC,WAAW,EAAE,IAAI,CAAC,WAAW,EAAE,MAAM,CAAC,EACvD,eAAe,EAAE,IAAI,CAAC,eAAe,EAAE,YAAY,CAAC;IAKtD;;;;;;;OAOG;IACG,WAAW,CAAC,OAAO,EAAE,YAAY,GAAG,OAAO,CAAC,UAAU,CAAC;IAoD7D,uCAAuC;IACvC,OAAO,CAAC,aAAa;IAyBrB;;;OAGG;YACW,cAAc;IA2E5B,6CAA6C;IAC7C,OAAO,CAAC,iBAAiB;IA+BzB,+DAA+D;IAC/D,UAAU,IAAI,IAAI;CAInB"}