@revealui/ai 0.1.0

package/dist/llm/provider-health.js

@@ -0,0 +1,97 @@
+/**
+ * Provider Health Monitor
+ *
+ * Sliding window of last 100 calls per provider.
+ * Tracks latency (p50) and error rate to rank fallback candidates.
+ *
+ * AnythingLLM lesson: no health checks → silent failures cascade.
+ */
+const WINDOW_SIZE = 100;
+const DEGRADED_ERROR_RATE = 0.1; // 10%
+const UNHEALTHY_ERROR_RATE = 0.3; // 30%
+const DEGRADED_LATENCY_MS = 5_000;
+const UNHEALTHY_LATENCY_MS = 15_000;
+function p50(values) {
+    if (values.length === 0)
+        return 0;
+    const sorted = [...values].sort((a, b) => a - b);
+    const mid = Math.floor(sorted.length / 2);
+    return sorted[mid] ?? 0;
+}
+export class ProviderHealthMonitor {
+    windows = new Map();
+    getWindow(provider) {
+        if (!this.windows.has(provider)) {
+            this.windows.set(provider, []);
+        }
+        // biome-ignore lint/style/noNonNullAssertion: set above guarantees presence
+        return this.windows.get(provider);
+    }
+    /**
+     * Record a completed LLM API call.
+     */
+    recordCall(provider, latencyMs, error) {
+        const window = this.getWindow(provider);
+        window.push({ latencyMs, error: error != null, timestamp: Date.now() });
+        // Keep only the last WINDOW_SIZE records
+        if (window.length > WINDOW_SIZE) {
+            window.splice(0, window.length - WINDOW_SIZE);
+        }
+    }
+    /**
+     * Get health metrics for a provider.
+     * Returns 'healthy' with 0 latency if no calls recorded yet.
+     */
+    getHealth(provider) {
+        const window = this.getWindow(provider);
+        if (window.length === 0) {
+            return {
+                provider,
+                status: 'healthy',
+                latencyP50Ms: 0,
+                errorRate: 0,
+                sampleCount: 0,
+            };
+        }
+        const errorCount = window.filter((r) => r.error).length;
+        const errorRate = errorCount / window.length;
+        const latencies = window.filter((r) => !r.error).map((r) => r.latencyMs);
+        const latencyP50Ms = p50(latencies);
+        let status = 'healthy';
+        if (errorRate >= UNHEALTHY_ERROR_RATE || latencyP50Ms >= UNHEALTHY_LATENCY_MS) {
+            status = 'unhealthy';
+        }
+        else if (errorRate >= DEGRADED_ERROR_RATE || latencyP50Ms >= DEGRADED_LATENCY_MS) {
+            status = 'degraded';
+        }
+        return { provider, status, latencyP50Ms, errorRate, sampleCount: window.length };
+    }
+    /**
+     * Pick the best provider from a set of candidates.
+     * Prefers healthy > degraded > unhealthy, then by p50 latency.
+     */
+    getBestProvider(candidates) {
+        if (candidates.length === 0)
+            throw new Error('No provider candidates provided');
+        // biome-ignore lint/style/noNonNullAssertion: length === 1 guarantees element exists
+        if (candidates.length === 1)
+            return candidates[0];
+        const ranked = candidates
+            .map((p) => ({ provider: p, health: this.getHealth(p) }))
+            .sort((a, b) => {
+            const statusOrder = { healthy: 0, degraded: 1, unhealthy: 2 };
+            const statusDiff = statusOrder[a.health.status] - statusOrder[b.health.status];
+            if (statusDiff !== 0)
+                return statusDiff;
+            return a.health.latencyP50Ms - b.health.latencyP50Ms;
+        });
+        // biome-ignore lint/style/noNonNullAssertion: ranked has at least 2 elements (checked above)
+        return ranked[0].provider;
+    }
+    /**
+     * Reset health data for a provider (useful for testing).
+     */
+    reset(provider) {
+        this.windows.set(provider, []);
+    }
+}

package/dist/llm/providers/anthropic.d.ts

@@ -0,0 +1,31 @@
+/**
+ * Anthropic Provider
+ *
+ * Implementation of LLMProvider for Anthropic Claude API
+ */
+import type { Embedding, LLMChatOptions, LLMChunk, LLMEmbedOptions, LLMProvider, LLMProviderConfig, LLMResponse, LLMStreamOptions, Message } from './base.js';
+export interface AnthropicProviderConfig extends LLMProviderConfig {
+    apiVersion?: string;
+    /** Enable prompt caching by default (5min TTL, 90% cost reduction on cache hits) */
+    enableCacheByDefault?: boolean;
+}
+export declare class AnthropicProvider implements LLMProvider {
+    private config;
+    private baseURL;
+    constructor(config: AnthropicProviderConfig);
+    chat(messages: Message[], options?: LLMChatOptions): Promise<LLMResponse>;
+    embed(text: string | string[], options?: LLMEmbedOptions): Promise<Embedding | Embedding[]>;
+    stream(messages: Message[], options?: LLMStreamOptions): AsyncIterable<LLMChunk>;
+    /**
+     * Format system messages with optional caching
+     * Caches the last system message for maximum benefit
+     */
+    private formatSystemMessages;
+    /**
+     * Format tools with optional caching
+     * Caches the last tool definition for maximum benefit
+     */
+    private formatTools;
+    private formatMessages;
+}
+//# sourceMappingURL=anthropic.d.ts.map

package/dist/llm/providers/anthropic.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"anthropic.d.ts","sourceRoot":"","sources":["../../../src/llm/providers/anthropic.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,KAAK,EACV,SAAS,EACT,cAAc,EACd,QAAQ,EACR,eAAe,EACf,WAAW,EACX,iBAAiB,EACjB,WAAW,EACX,gBAAgB,EAChB,OAAO,EAER,MAAM,WAAW,CAAA;AAElB,MAAM,WAAW,uBAAwB,SAAQ,iBAAiB;IAChE,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,oFAAoF;IACpF,oBAAoB,CAAC,EAAE,OAAO,CAAA;CAC/B;AAqDD,qBAAa,iBAAkB,YAAW,WAAW;IACnD,OAAO,CAAC,MAAM,CAAyB;IACvC,OAAO,CAAC,OAAO,CAAQ;gBAEX,MAAM,EAAE,uBAAuB;IAKrC,IAAI,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC;IA6F/E,KAAK,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,EAAE,OAAO,CAAC,EAAE,eAAe,GAAG,OAAO,CAAC,SAAS,GAAG,SAAS,EAAE,CAAC;IAUpF,MAAM,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,aAAa,CAAC,QAAQ,CAAC;IAoGvF;;;OAGG;IACH,OAAO,CAAC,oBAAoB;IA0B5B;;;OAGG;IACH,OAAO,CAAC,WAAW;IAmBnB,OAAO,CAAC,cAAc;CAiBvB"}

package/dist/llm/providers/anthropic.js

@@ -0,0 +1,248 @@
+/**
+ * Anthropic Provider
+ *
+ * Implementation of LLMProvider for Anthropic Claude API
+ */
+const isRecord = (value) => typeof value === 'object' && value !== null && !Array.isArray(value);
+const isTextBlock = (block) => block.type === 'text' && typeof block.text === 'string';
+const isToolUseBlock = (block) => block.type === 'tool_use';
+const maxTokensKey = 'max_tokens';
+const inputTokensKey = 'input_tokens';
+const outputTokensKey = 'output_tokens';
+const stopReasonKey = 'stop_reason';
+const cacheCreationTokensKey = 'cache_creation_input_tokens';
+const cacheReadTokensKey = 'cache_read_input_tokens';
+export class AnthropicProvider {
+    config;
+    baseURL;
+    constructor(config) {
+        this.config = config;
+        this.baseURL = config.baseURL || 'https://api.anthropic.com/v1';
+    }
+    async chat(messages, options) {
+        // Anthropic API format is slightly different
+        const systemMessages = messages.filter((m) => m.role === 'system');
+        const conversationMessages = messages.filter((m) => m.role !== 'system');
+        const enableCache = options?.enableCache ?? this.config.enableCacheByDefault ?? false;
+        // Use 2024-07-15 API version for prompt caching support
+        const apiVersion = enableCache ? '2024-07-15' : this.config.apiVersion || '2023-06-01';
+        // Format system messages with caching
+        const systemContent = this.formatSystemMessages(systemMessages, enableCache);
+        // Format tools with caching (cache last tool if enabled)
+        const tools = this.formatTools(options?.tools, enableCache);
+        const response = await fetch(`${this.baseURL}/messages`, {
+            method: 'POST',
+            headers: {
+                'Content-Type': 'application/json',
+                'x-api-key': this.config.apiKey,
+                'anthropic-version': apiVersion,
+            },
+            body: JSON.stringify({
+                model: this.config.model || 'claude-3-5-sonnet-20241022',
+                system: systemContent,
+                messages: this.formatMessages(conversationMessages),
+                temperature: options?.temperature ?? this.config.temperature ?? 0.7,
+                [maxTokensKey]: options?.maxTokens ?? this.config.maxTokens ?? 4096,
+                tools,
+            }),
+        });
+        if (!response.ok) {
+            const errorPayload = (await response.json().catch(() => undefined));
+            const errorMessage = isRecord(errorPayload) &&
+                isRecord(errorPayload.error) &&
+                typeof errorPayload.error.message === 'string'
+                ? errorPayload.error.message
+                : response.statusText;
+            throw new Error(`Anthropic API error: ${errorMessage}`);
+        }
+        const data = (await response.json());
+        const contentBlocks = Array.isArray(data.content)
+            ? data.content
+            : [];
+        const textBlock = contentBlocks.find(isTextBlock);
+        const toolCalls = contentBlocks.filter(isToolUseBlock).map((tc) => ({
+            id: tc.id,
+            type: 'function',
+            function: {
+                name: tc.name,
+                arguments: JSON.stringify(tc.input),
+            },
+        }));
+        const usage = data.usage && typeof data.usage === 'object'
+            ? data.usage
+            : undefined;
+        const inputTokens = usage && typeof usage[inputTokensKey] === 'number' ? usage[inputTokensKey] : undefined;
+        const outputTokens = usage && typeof usage[outputTokensKey] === 'number' ? usage[outputTokensKey] : undefined;
+        const cacheCreationTokens = usage && typeof usage[cacheCreationTokensKey] === 'number'
+            ? usage[cacheCreationTokensKey]
+            : undefined;
+        const cacheReadTokens = usage && typeof usage[cacheReadTokensKey] === 'number' ? usage[cacheReadTokensKey] : undefined;
+        const finishReason = typeof data[stopReasonKey] === 'string'
+            ? data[stopReasonKey]
+            : undefined;
+        return {
+            content: textBlock?.text || '',
+            role: 'assistant',
+            toolCalls,
+            finishReason,
+            usage: inputTokens !== undefined && outputTokens !== undefined
+                ? {
+                    promptTokens: inputTokens,
+                    completionTokens: outputTokens,
+                    totalTokens: inputTokens + outputTokens,
+                    cacheCreationTokens,
+                    cacheReadTokens,
+                }
+                : undefined,
+        };
+    }
+    embed(text, options) {
+        void text;
+        void options;
+        // Anthropic doesn't have a separate embeddings API
+        // Would need to use a different provider or service
+        return Promise.reject(new Error('Anthropic does not support embeddings. Use OpenAI provider for embeddings.'));
+    }
+    async *stream(messages, options) {
+        const systemMessages = messages.filter((m) => m.role === 'system');
+        const conversationMessages = messages.filter((m) => m.role !== 'system');
+        const enableCache = options?.enableCache ?? this.config.enableCacheByDefault ?? false;
+        // Use 2024-07-15 API version for prompt caching support
+        const apiVersion = enableCache ? '2024-07-15' : this.config.apiVersion || '2023-06-01';
+        // Format system messages with caching
+        const systemContent = this.formatSystemMessages(systemMessages, enableCache);
+        // Format tools with caching
+        const tools = this.formatTools(options?.tools, enableCache);
+        const response = await fetch(`${this.baseURL}/messages`, {
+            method: 'POST',
+            headers: {
+                'Content-Type': 'application/json',
+                'x-api-key': this.config.apiKey,
+                'anthropic-version': apiVersion,
+            },
+            body: JSON.stringify({
+                model: this.config.model || 'claude-3-5-sonnet-20241022',
+                system: systemContent,
+                messages: this.formatMessages(conversationMessages),
+                temperature: options?.temperature ?? this.config.temperature ?? 0.7,
+                [maxTokensKey]: options?.maxTokens ?? this.config.maxTokens ?? 4096,
+                tools,
+                stream: true,
+            }),
+        });
+        if (!response.ok) {
+            const errorPayload = (await response.json().catch(() => undefined));
+            const errorMessage = isRecord(errorPayload) &&
+                isRecord(errorPayload.error) &&
+                typeof errorPayload.error.message === 'string'
+                ? errorPayload.error.message
+                : response.statusText;
+            throw new Error(`Anthropic API error: ${errorMessage}`);
+        }
+        const reader = response.body?.getReader();
+        const decoder = new TextDecoder();
+        if (!reader) {
+            throw new Error('Response body is not readable');
+        }
+        let buffer = '';
+        while (true) {
+            const { done, value } = await reader.read();
+            if (done) {
+                yield { content: '', done: true };
+                break;
+            }
+            buffer += decoder.decode(value, { stream: true });
+            const lines = buffer.split('\n');
+            buffer = lines.pop() || '';
+            for (const line of lines) {
+                if (line.startsWith('data: ')) {
+                    const data = line.slice(6);
+                    if (data === '[DONE]') {
+                        yield { content: '', done: true };
+                        return;
+                    }
+                    try {
+                        const parsed = JSON.parse(data);
+                        if (!isRecord(parsed)) {
+                            continue;
+                        }
+                        const eventType = typeof parsed.type === 'string' ? parsed.type : undefined;
+                        if (eventType === 'content_block_delta' && isRecord(parsed.delta)) {
+                            const deltaType = typeof parsed.delta.type === 'string' ? parsed.delta.type : undefined;
+                            if (deltaType === 'text_delta') {
+                                yield {
+                                    content: typeof parsed.delta.text === 'string' ? parsed.delta.text : '',
+                                    done: false,
+                                };
+                            }
+                        }
+                        else if (eventType === 'message_stop') {
+                            yield { content: '', done: true };
+                            return;
+                        }
+                    }
+                    catch {
+                        // Ignore parse errors for incomplete chunks
+                    }
+                }
+            }
+        }
+    }
+    /**
+     * Format system messages with optional caching
+     * Caches the last system message for maximum benefit
+     */
+    formatSystemMessages(systemMessages, enableCache) {
+        if (systemMessages.length === 0) {
+            return '';
+        }
+        // If caching disabled, use simple string format
+        if (!enableCache) {
+            return systemMessages.map((m) => m.content).join('\n');
+        }
+        // With caching, use structured format and cache the last block
+        return systemMessages.map((msg, index) => ({
+            type: 'text',
+            text: msg.content,
+            // Cache the last system message (most likely to be reused)
+            ...(index === systemMessages.length - 1 && msg.cacheControl
+                ? { cache_control: msg.cacheControl }
+                : index === systemMessages.length - 1
+                    ? { cache_control: { type: 'ephemeral' } }
+                    : {}),
+        }));
+    }
+    /**
+     * Format tools with optional caching
+     * Caches the last tool definition for maximum benefit
+     */
+    formatTools(tools, enableCache) {
+        if (!tools || tools.length === 0) {
+            return undefined;
+        }
+        return tools.map((tool, index) => ({
+            name: tool.function.name,
+            description: tool.function.description,
+            input_schema: tool.function.parameters,
+            // Cache the last tool (most likely to be reused across calls)
+            ...(enableCache && index === tools.length - 1
+                ? { cache_control: { type: 'ephemeral' } }
+                : {}),
+        }));
+    }
+    formatMessages(messages) {
+        return messages
+            .map((msg) => {
+            if (msg.role === 'system') {
+                // System messages are handled separately in Anthropic API
+                return null;
+            }
+            const formatted = {
+                role: msg.role === 'assistant' ? 'assistant' : 'user',
+                content: msg.content,
+            };
+            return formatted;
+        })
+            .filter((message) => Boolean(message));
+    }
+}

package/dist/llm/providers/base.d.ts

@@ -0,0 +1,111 @@
+/**
+ * LLM Provider Base Interface
+ *
+ * Abstract interface for all LLM providers (OpenAI, Anthropic, etc.)
+ */
+export interface Message {
+    role: 'system' | 'user' | 'assistant' | 'tool';
+    content: string;
+    name?: string;
+    toolCalls?: ToolCall[];
+    toolCallId?: string;
+    /** Anthropic prompt caching - marks content for caching (5min TTL, 90% cost reduction) */
+    cacheControl?: {
+        type: 'ephemeral';
+    };
+}
+export interface ToolCall {
+    id: string;
+    type: 'function';
+    function: {
+        name: string;
+        arguments: string;
+    };
+}
+export interface LLMResponse {
+    content: string;
+    role: 'assistant';
+    toolCalls?: ToolCall[];
+    finishReason?: 'stop' | 'length' | 'tool_calls' | 'content_filter';
+    usage?: {
+        promptTokens: number;
+        completionTokens: number;
+        totalTokens: number;
+        /** Anthropic cache stats */
+        cacheCreationTokens?: number;
+        cacheReadTokens?: number;
+    };
+}
+export interface Embedding {
+    vector: number[];
+    dimension: number;
+    model: string;
+}
+export interface LLMChunk {
+    content: string;
+    done: boolean;
+    toolCalls?: ToolCall[];
+}
+export interface LLMProviderConfig {
+    apiKey: string;
+    baseURL?: string;
+    model?: string;
+    temperature?: number;
+    maxTokens?: number;
+}
+/**
+ * Base interface for LLM providers
+ */
+export interface LLMProvider {
+    /**
+     * Chat completion
+     */
+    chat(messages: Message[], options?: LLMChatOptions): Promise<LLMResponse>;
+    /**
+     * Generate embeddings
+     */
+    embed(text: string | string[], options?: LLMEmbedOptions): Promise<Embedding | Embedding[]>;
+    /**
+     * Stream chat completion
+     */
+    stream(messages: Message[], options?: LLMStreamOptions): AsyncIterable<LLMChunk>;
+}
+export interface LLMChatOptions {
+    temperature?: number;
+    maxTokens?: number;
+    tools?: ToolDefinition[];
+    toolChoice?: 'auto' | 'none' | {
+        type: 'function';
+        function: {
+            name: string;
+        };
+    };
+    /** Enable prompt caching (Anthropic only) - caches system prompts and tools */
+    enableCache?: boolean;
+    /**
+     * Extended thinking token budget (Anthropic only).
+     * Enables Claude's internal reasoning before responding.
+     * Typical values: 512 (minimal) → 31999 (xhigh). 0 or undefined = disabled.
+     */
+    thinkingBudget?: number;
+}
+export interface LLMEmbedOptions {
+    model?: string;
+}
+export interface LLMStreamOptions {
+    temperature?: number;
+    maxTokens?: number;
+    tools?: ToolDefinition[];
+    /** Enable prompt caching (Anthropic only) - caches system prompts and tools */
+    enableCache?: boolean;
+}
+export interface ToolDefinition {
+    type: 'function';
+    function: {
+        name: string;
+        description: string;
+        parameters: Record<string, unknown>;
+    };
+}
+export type FinishReason = 'stop' | 'length' | 'tool_calls' | 'content_filter';
+//# sourceMappingURL=base.d.ts.map

package/dist/llm/providers/base.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"base.d.ts","sourceRoot":"","sources":["../../../src/llm/providers/base.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,MAAM,WAAW,OAAO;IACtB,IAAI,EAAE,QAAQ,GAAG,MAAM,GAAG,WAAW,GAAG,MAAM,CAAA;IAC9C,OAAO,EAAE,MAAM,CAAA;IACf,IAAI,CAAC,EAAE,MAAM,CAAA;IACb,SAAS,CAAC,EAAE,QAAQ,EAAE,CAAA;IACtB,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,0FAA0F;IAC1F,YAAY,CAAC,EAAE;QAAE,IAAI,EAAE,WAAW,CAAA;KAAE,CAAA;CACrC;AAED,MAAM,WAAW,QAAQ;IACvB,EAAE,EAAE,MAAM,CAAA;IACV,IAAI,EAAE,UAAU,CAAA;IAChB,QAAQ,EAAE;QACR,IAAI,EAAE,MAAM,CAAA;QACZ,SAAS,EAAE,MAAM,CAAA;KAClB,CAAA;CACF;AAED,MAAM,WAAW,WAAW;IAC1B,OAAO,EAAE,MAAM,CAAA;IACf,IAAI,EAAE,WAAW,CAAA;IACjB,SAAS,CAAC,EAAE,QAAQ,EAAE,CAAA;IACtB,YAAY,CAAC,EAAE,MAAM,GAAG,QAAQ,GAAG,YAAY,GAAG,gBAAgB,CAAA;IAClE,KAAK,CAAC,EAAE;QACN,YAAY,EAAE,MAAM,CAAA;QACpB,gBAAgB,EAAE,MAAM,CAAA;QACxB,WAAW,EAAE,MAAM,CAAA;QACnB,4BAA4B;QAC5B,mBAAmB,CAAC,EAAE,MAAM,CAAA;QAC5B,eAAe,CAAC,EAAE,MAAM,CAAA;KACzB,CAAA;CACF;AAED,MAAM,WAAW,SAAS;IACxB,MAAM,EAAE,MAAM,EAAE,CAAA;IAChB,SAAS,EAAE,MAAM,CAAA;IACjB,KAAK,EAAE,MAAM,CAAA;CACd;AAED,MAAM,WAAW,QAAQ;IACvB,OAAO,EAAE,MAAM,CAAA;IACf,IAAI,EAAE,OAAO,CAAA;IACb,SAAS,CAAC,EAAE,QAAQ,EAAE,CAAA;CACvB;AAED,MAAM,WAAW,iBAAiB;IAChC,MAAM,EAAE,MAAM,CAAA;IACd,OAAO,CAAC,EAAE,MAAM,CAAA;IAChB,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,SAAS,CAAC,EAAE,MAAM,CAAA;CACnB;AAED;;GAEG;AACH,MAAM,WAAW,WAAW;IAC1B;;OAEG;IACH,IAAI,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC,CAAA;IAEzE;;OAEG;IACH,KAAK,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,EAAE,OAAO,CAAC,EAAE,eAAe,GAAG,OAAO,CAAC,SAAS,GAAG,SAAS,EAAE,CAAC,CAAA;IAE3F;;OAEG;IACH,MAAM,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,aAAa,CAAC,QAAQ,CAAC,CAAA;CACjF;AAED,MAAM,WAAW,cAAc;IAC7B,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,KAAK,CAAC,EAAE,cAAc,EAAE,CAAA;IACxB,UAAU,CAAC,EAAE,MAAM,GAAG,MAAM,GAAG;QAAE,IAAI,EAAE,UAAU,CAAC;QAAC,QAAQ,EAAE;YAAE,IAAI,EAAE,MAAM,CAAA;SAAE,CAAA;KAAE,CAAA;IAC/E,+EAA+E;IAC/E,WAAW,CAAC,EAAE,OAAO,CAAA;IACrB;;;;OAIG;IACH,cAAc,CAAC,EAAE,MAAM,CAAA;CACxB;AAED,MAAM,WAAW,eAAe;IAC9B,KAAK,CAAC,EAAE,MAAM,CAAA;CACf;AAED,MAAM,WAAW,gBAAgB;IAC/B,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,KAAK,CAAC,EAAE,cAAc,EAAE,CAAA;IACxB,+EAA+E;IAC/E,WAAW,CAAC,EAAE,OAAO,CAAA;CACtB;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE,UAAU,CAAA;IAChB,QAAQ,EAAE;QACR,IAAI,EAAE,MAAM,CAAA;QACZ,WAAW,EAAE,MAAM,CAAA;QACnB,UAAU,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAA;KACpC,CAAA;CACF;AAED,MAAM,MAAM,YAAY,GAAG,MAAM,GAAG,QAAQ,GAAG,YAAY,GAAG,gBAAgB,CAAA"}

package/dist/llm/providers/base.js

@@ -0,0 +1,6 @@
+/**
+ * LLM Provider Base Interface
+ *
+ * Abstract interface for all LLM providers (OpenAI, Anthropic, etc.)
+ */
+export {};

package/dist/llm/providers/groq.d.ts

@@ -0,0 +1,23 @@
+/**
+ * Groq Provider
+ *
+ * Thin wrapper over OpenAIProvider using Groq's OpenAI-compatible API.
+ * Free tier: 6,000 TPM / 500k TPD for Llama 3.3 70B.
+ * Sign up: console.groq.com
+ */
+import type { Embedding, LLMChatOptions, LLMChunk, LLMEmbedOptions, LLMProvider, LLMProviderConfig, LLMResponse, LLMStreamOptions, Message } from './base.js';
+export interface GroqProviderConfig extends Omit<LLMProviderConfig, 'apiKey'> {
+    apiKey: string;
+    /** Defaults to https://api.groq.com/openai/v1 */
+    baseURL?: string;
+    /** Defaults to llama-3.3-70b-versatile */
+    model?: string;
+}
+export declare class GroqProvider implements LLMProvider {
+    private inner;
+    constructor(config: GroqProviderConfig);
+    chat(messages: Message[], options?: LLMChatOptions): Promise<LLMResponse>;
+    stream(messages: Message[], options?: LLMStreamOptions): AsyncIterable<LLMChunk>;
+    embed(_text: string | string[], _options?: LLMEmbedOptions): Promise<Embedding | Embedding[]>;
+}
+//# sourceMappingURL=groq.d.ts.map

package/dist/llm/providers/groq.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"groq.d.ts","sourceRoot":"","sources":["../../../src/llm/providers/groq.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH,OAAO,KAAK,EACV,SAAS,EACT,cAAc,EACd,QAAQ,EACR,eAAe,EACf,WAAW,EACX,iBAAiB,EACjB,WAAW,EACX,gBAAgB,EAChB,OAAO,EACR,MAAM,WAAW,CAAA;AAGlB,MAAM,WAAW,kBAAmB,SAAQ,IAAI,CAAC,iBAAiB,EAAE,QAAQ,CAAC;IAC3E,MAAM,EAAE,MAAM,CAAA;IACd,iDAAiD;IACjD,OAAO,CAAC,EAAE,MAAM,CAAA;IAChB,0CAA0C;IAC1C,KAAK,CAAC,EAAE,MAAM,CAAA;CACf;AAED,qBAAa,YAAa,YAAW,WAAW;IAC9C,OAAO,CAAC,KAAK,CAAgB;gBAEjB,MAAM,EAAE,kBAAkB;IAQtC,IAAI,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC;IAIzE,MAAM,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,aAAa,CAAC,QAAQ,CAAC;IAIhF,KAAK,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,EAAE,EAAE,QAAQ,CAAC,EAAE,eAAe,GAAG,OAAO,CAAC,SAAS,GAAG,SAAS,EAAE,CAAC;CAG9F"}

package/dist/llm/providers/groq.js

@@ -0,0 +1,27 @@
+/**
+ * Groq Provider
+ *
+ * Thin wrapper over OpenAIProvider using Groq's OpenAI-compatible API.
+ * Free tier: 6,000 TPM / 500k TPD for Llama 3.3 70B.
+ * Sign up: console.groq.com
+ */
+import { OpenAIProvider } from './openai.js';
+export class GroqProvider {
+    inner;
+    constructor(config) {
+        this.inner = new OpenAIProvider({
+            ...config,
+            baseURL: config.baseURL ?? 'https://api.groq.com/openai/v1',
+            model: config.model ?? 'llama-3.3-70b-versatile',
+        });
+    }
+    chat(messages, options) {
+        return this.inner.chat(messages, options);
+    }
+    stream(messages, options) {
+        return this.inner.stream(messages, options);
+    }
+    embed(_text, _options) {
+        throw new Error('Groq does not support embeddings. Use Ollama or HuggingFace.');
+    }
+}

package/dist/llm/providers/ollama.d.ts

@@ -0,0 +1,27 @@
+/**
+ * Ollama Provider
+ *
+ * Local inference via Ollama's OpenAI-compatible API (http://localhost:11434/v1).
+ * No API key required. Zero cost, fully offline.
+ * Install: https://ollama.com
+ */
+import type { Embedding, LLMChatOptions, LLMChunk, LLMEmbedOptions, LLMProvider, LLMProviderConfig, LLMResponse, LLMStreamOptions, Message } from './base.js';
+export interface OllamaProviderConfig extends Omit<LLMProviderConfig, 'apiKey'> {
+    apiKey?: string;
+    /** Defaults to http://localhost:11434/v1 */
+    baseURL?: string;
+    /** Chat model. Defaults to llama3.2:3b — run `ollama pull llama3.2:3b` first */
+    model?: string;
+    /** Embedding model. Defaults to nomic-embed-text — run `ollama pull nomic-embed-text` first */
+    embedModel?: string;
+}
+export declare class OllamaProvider implements LLMProvider {
+    private inner;
+    private embedModel;
+    private baseURL;
+    constructor(config: OllamaProviderConfig);
+    chat(messages: Message[], options?: LLMChatOptions): Promise<LLMResponse>;
+    stream(messages: Message[], options?: LLMStreamOptions): AsyncIterable<LLMChunk>;
+    embed(text: string | string[], _options?: LLMEmbedOptions): Promise<Embedding | Embedding[]>;
+}
+//# sourceMappingURL=ollama.d.ts.map

package/dist/llm/providers/ollama.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"ollama.d.ts","sourceRoot":"","sources":["../../../src/llm/providers/ollama.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AAEH,OAAO,KAAK,EACV,SAAS,EACT,cAAc,EACd,QAAQ,EACR,eAAe,EACf,WAAW,EACX,iBAAiB,EACjB,WAAW,EACX,gBAAgB,EAChB,OAAO,EACR,MAAM,WAAW,CAAA;AAGlB,MAAM,WAAW,oBAAqB,SAAQ,IAAI,CAAC,iBAAiB,EAAE,QAAQ,CAAC;IAC7E,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,4CAA4C;IAC5C,OAAO,CAAC,EAAE,MAAM,CAAA;IAChB,gFAAgF;IAChF,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,+FAA+F;IAC/F,UAAU,CAAC,EAAE,MAAM,CAAA;CACpB;AAED,qBAAa,cAAe,YAAW,WAAW;IAChD,OAAO,CAAC,KAAK,CAAgB;IAC7B,OAAO,CAAC,UAAU,CAAQ;IAC1B,OAAO,CAAC,OAAO,CAAQ;gBAEX,MAAM,EAAE,oBAAoB;IAaxC,IAAI,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC;IAIzE,MAAM,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,aAAa,CAAC,QAAQ,CAAC;IAI1E,KAAK,CACT,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,EACvB,QAAQ,CAAC,EAAE,eAAe,GACzB,OAAO,CAAC,SAAS,GAAG,SAAS,EAAE,CAAC;CAsBpC"}

package/dist/llm/providers/ollama.js

@@ -0,0 +1,48 @@
+/**
+ * Ollama Provider
+ *
+ * Local inference via Ollama's OpenAI-compatible API (http://localhost:11434/v1).
+ * No API key required. Zero cost, fully offline.
+ * Install: https://ollama.com
+ */
+import { OpenAIProvider } from './openai.js';
+export class OllamaProvider {
+    inner;
+    embedModel;
+    baseURL;
+    constructor(config) {
+        const baseURL = config.baseURL ?? 'http://localhost:11434/v1';
+        this.baseURL = baseURL;
+        this.embedModel = config.embedModel ?? 'nomic-embed-text';
+        this.inner = new OpenAIProvider({
+            ...config,
+            // Ollama ignores the API key but the OpenAI client requires a non-empty value
+            apiKey: config.apiKey ?? 'ollama',
+            baseURL,
+            model: config.model ?? 'llama3.2:3b',
+        });
+    }
+    chat(messages, options) {
+        return this.inner.chat(messages, options);
+    }
+    stream(messages, options) {
+        return this.inner.stream(messages, options);
+    }
+    async embed(text, _options) {
+        const texts = Array.isArray(text) ? text : [text];
+        const response = await fetch(`${this.baseURL}/embeddings`, {
+            method: 'POST',
+            headers: { 'Content-Type': 'application/json' },
+            body: JSON.stringify({ model: this.embedModel, input: texts }),
+        });
+        if (!response.ok) {
+            throw new Error(`Ollama embeddings error: ${response.statusText}`);
+        }
+        const data = (await response.json());
+        const embeddings = (data.data ?? []).map((item) => {
+            const vector = item.embedding ?? [];
+            return { vector, dimension: vector.length, model: this.embedModel };
+        });
+        return Array.isArray(text) ? embeddings : embeddings[0];
+    }
+}

package/dist/llm/providers/openai.d.ts

@@ -0,0 +1,19 @@
+/**
+ * OpenAI Provider
+ *
+ * Implementation of LLMProvider for OpenAI API
+ */
+import type { Embedding, LLMChatOptions, LLMChunk, LLMEmbedOptions, LLMProvider, LLMProviderConfig, LLMResponse, LLMStreamOptions, Message } from './base.js';
+export interface OpenAIProviderConfig extends LLMProviderConfig {
+    organization?: string;
+}
+export declare class OpenAIProvider implements LLMProvider {
+    private config;
+    private baseURL;
+    constructor(config: OpenAIProviderConfig);
+    chat(messages: Message[], options?: LLMChatOptions): Promise<LLMResponse>;
+    embed(text: string | string[], options?: LLMEmbedOptions): Promise<Embedding | Embedding[]>;
+    stream(messages: Message[], options?: LLMStreamOptions): AsyncIterable<LLMChunk>;
+    private formatMessages;
+}
+//# sourceMappingURL=openai.d.ts.map

package/dist/llm/providers/openai.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"openai.d.ts","sourceRoot":"","sources":["../../../src/llm/providers/openai.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAEH,OAAO,KAAK,EACV,SAAS,EAET,cAAc,EACd,QAAQ,EACR,eAAe,EACf,WAAW,EACX,iBAAiB,EACjB,WAAW,EACX,gBAAgB,EAChB,OAAO,EAER,MAAM,WAAW,CAAA;AAElB,MAAM,WAAW,oBAAqB,SAAQ,iBAAiB;IAC7D,YAAY,CAAC,EAAE,MAAM,CAAA;CACtB;AAsDD,qBAAa,cAAe,YAAW,WAAW;IAChD,OAAO,CAAC,MAAM,CAAsB;IACpC,OAAO,CAAC,OAAO,CAAQ;gBAEX,MAAM,EAAE,oBAAoB;IAKlC,IAAI,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC;IAgFzE,KAAK,CACT,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,EACvB,OAAO,CAAC,EAAE,eAAe,GACxB,OAAO,CAAC,SAAS,GAAG,SAAS,EAAE,CAAC;IA2C5B,MAAM,CAAC,QAAQ,EAAE,OAAO,EAAE,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,aAAa,CAAC,QAAQ,CAAC;IA0FvF,OAAO,CAAC,cAAc;CA0BvB"}