@molroo-io/sdk 0.6.2 → 0.7.1

package/dist/cjs/memory/supabase/memory-adapter.js

@@ -1,335 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.SupabaseMemoryAdapter = void 0;
-const supabase_js_1 = require("@supabase/supabase-js");
-/**
- * Supabase-backed MemoryAdapter with optional pgvector semantic search.
- *
- * Combines episode/reflection RDB storage with optional pgvector embeddings
- * into a single unified MemoryAdapter interface.
- *
- * @example
- * ```typescript
- * // RDB only (no vector search)
- * const adapter = new SupabaseMemoryAdapter({
- *   supabaseUrl: process.env.SUPABASE_URL!,
- *   supabaseKey: process.env.SUPABASE_KEY!,
- *   personaId: 'world-123',
- * });
- *
- * // With vector search
- * const adapter = new SupabaseMemoryAdapter({
- *   supabaseUrl: process.env.SUPABASE_URL!,
- *   supabaseKey: process.env.SUPABASE_KEY!,
- *   personaId: 'world-123',
- *   vector: {
- *     embedding: openaiEmbedding({ apiKey: '...' }),
- *     matchThreshold: 0.75,
- *   },
- * });
- * ```
- */
-class SupabaseMemoryAdapter {
-    constructor(options) {
-        this.client = (0, supabase_js_1.createClient)(options.supabaseUrl, options.supabaseKey);
-        this.personaId = options.personaId;
-        this.rlsStrategy = options.rlsStrategy ?? 'application';
-        this.vectorEnabled = !!options.vector;
-        this.embedding = options.vector?.embedding;
-        this.matchThreshold = options.vector?.matchThreshold ?? 0.7;
-    }
-    async setPersonaContext() {
-        if (this.rlsStrategy === 'application') {
-            await this.client
-                .rpc('set_config', {
-                setting: 'app.current_persona_id',
-                value: this.personaId,
-            })
-                .throwOnError();
-        }
-    }
-    // ── Required methods ──
-    async saveEpisode(episode) {
-        await this.setPersonaContext();
-        const { error } = await this.client.from('episodes').upsert({
-            id: episode.id,
-            persona_id: this.personaId,
-            source_entity: episode.sourceEntity ?? null,
-            context: episode.context ?? null,
-            timestamp: episode.timestamp,
-            importance: episode.importance,
-            intensity: episode.intensity,
-            vad_v: episode.emotionSnapshot.V,
-            vad_a: episode.emotionSnapshot.A,
-            vad_d: episode.emotionSnapshot.D,
-            appraisal: episode.appraisal,
-        });
-        if (error)
-            throw new Error(`saveEpisode failed: ${error.message}`);
-        // Auto-index embedding when vector is enabled
-        if (this.vectorEnabled && this.embedding) {
-            try {
-                const text = episode.context ?? episode.sourceEntity ?? '';
-                if (text) {
-                    const embedding = await this.embedding.embed(text);
-                    await this.indexEmbedding(episode.id, embedding, {
-                        type: 'episode',
-                        sourceEntity: episode.sourceEntity,
-                        timestamp: episode.timestamp,
-                        importance: episode.importance,
-                    });
-                }
-            }
-            catch {
-                // Best-effort: vector indexing failure doesn't block episode save.
-            }
-        }
-    }
-    async recall(query) {
-        await this.setPersonaContext();
-        const limit = query.limit ?? 10;
-        let q = this.client
-            .from('episodes')
-            .select('*')
-            .eq('persona_id', this.personaId);
-        if (query.sourceEntity)
-            q = q.eq('source_entity', query.sourceEntity);
-        if (query.context)
-            q = q.ilike('context', `%${query.context}%`);
-        if (query.minImportance !== undefined)
-            q = q.gte('importance', query.minImportance);
-        if (query.valenceRange) {
-            q = q
-                .gte('vad_v', query.valenceRange[0])
-                .lte('vad_v', query.valenceRange[1]);
-        }
-        if (query.timeRange) {
-            q = q
-                .gte('timestamp', query.timeRange[0])
-                .lte('timestamp', query.timeRange[1]);
-        }
-        const { data, error } = await q
-            .order('timestamp', { ascending: false })
-            .limit(limit);
-        if (error)
-            throw new Error(`recall failed: ${error.message}`);
-        return (data ?? []).map(rowToEpisode);
-    }
-    async saveReflection(reflection) {
-        await this.setPersonaContext();
-        const { error } = await this.client.from('reflections').upsert({
-            id: reflection.id,
-            persona_id: this.personaId,
-            source_entity: reflection.sourceEntity ?? null,
-            content: reflection.content,
-            trigger: reflection.trigger,
-            timestamp: reflection.timestamp,
-            vad_v: reflection.emotionSnapshot.V,
-            vad_a: reflection.emotionSnapshot.A,
-            vad_d: reflection.emotionSnapshot.D,
-        });
-        if (error)
-            throw new Error(`saveReflection failed: ${error.message}`);
-    }
-    async getReflections(sourceEntity) {
-        await this.setPersonaContext();
-        const maxRows = 100;
-        let q = this.client
-            .from('reflections')
-            .select('*')
-            .eq('persona_id', this.personaId)
-            .order('timestamp', { ascending: false });
-        if (sourceEntity)
-            q = q.eq('source_entity', sourceEntity);
-        const { data, error } = await q.limit(maxRows);
-        if (error)
-            throw new Error(`getReflections failed: ${error.message}`);
-        return (data ?? []).map(rowToReflection);
-    }
-    // ── Optional methods ──
-    async semanticRecall(query, options) {
-        if (!this.vectorEnabled || !this.embedding) {
-            // Fallback to text-based recall
-            return this.recall({ context: query, limit: options?.topK ?? 10 });
-        }
-        await this.setPersonaContext();
-        const embedding = await this.embedding.embed(query);
-        const topK = options?.topK ?? 5;
-        const { data, error } = await this.client.rpc('match_embeddings', {
-            query_embedding: embedding,
-            match_threshold: this.matchThreshold,
-            match_count: topK,
-            filter_persona_id: this.personaId,
-            filter_type: 'episode',
-            filter_source_entity: options?.filter?.sourceEntity ?? null,
-            filter_min_importance: options?.filter?.minImportance ?? null,
-        });
-        if (error)
-            throw new Error(`semanticRecall failed: ${error.message}`);
-        const matchedIds = (data ?? []).map((row) => row.id);
-        if (matchedIds.length === 0)
-            return [];
-        // Fetch full episodes for matched IDs
-        return this.getByIds(matchedIds);
-    }
-    async getByIds(ids) {
-        if (ids.length === 0)
-            return [];
-        await this.setPersonaContext();
-        const { data, error } = await this.client
-            .from('episodes')
-            .select('*')
-            .eq('persona_id', this.personaId)
-            .in('id', ids);
-        if (error)
-            throw new Error(`getByIds failed: ${error.message}`);
-        return (data ?? []).map(rowToEpisode);
-    }
-    async countEpisodes(query) {
-        await this.setPersonaContext();
-        let q = this.client
-            .from('episodes')
-            .select('*', { count: 'exact', head: true })
-            .eq('persona_id', this.personaId);
-        if (query.sourceEntity)
-            q = q.eq('source_entity', query.sourceEntity);
-        if (query.context)
-            q = q.ilike('context', `%${query.context}%`);
-        if (query.timeRange) {
-            q = q
-                .gte('timestamp', query.timeRange[0])
-                .lte('timestamp', query.timeRange[1]);
-        }
-        const { count, error } = await q;
-        if (error)
-            throw new Error(`countEpisodes failed: ${error.message}`);
-        return count ?? 0;
-    }
-    async updateEpisode(id, patch) {
-        await this.setPersonaContext();
-        const updates = {};
-        if (patch.importance !== undefined)
-            updates.importance = patch.importance;
-        if (patch.intensity !== undefined)
-            updates.intensity = patch.intensity;
-        if (patch.emotionSnapshot !== undefined) {
-            updates.vad_v = patch.emotionSnapshot.V;
-            updates.vad_a = patch.emotionSnapshot.A;
-            updates.vad_d = patch.emotionSnapshot.D;
-        }
-        if (Object.keys(updates).length === 0)
-            return;
-        const { error } = await this.client
-            .from('episodes')
-            .update(updates)
-            .eq('id', id)
-            .eq('persona_id', this.personaId);
-        if (error)
-            throw new Error(`updateEpisode failed: ${error.message}`);
-    }
-    async deleteEpisode(id) {
-        await this.setPersonaContext();
-        const { error } = await this.client
-            .from('episodes')
-            .delete()
-            .eq('id', id)
-            .eq('persona_id', this.personaId);
-        if (error)
-            throw new Error(`deleteEpisode failed: ${error.message}`);
-        // Best-effort: clean up embedding
-        if (this.vectorEnabled) {
-            try {
-                await this.client
-                    .from('memory_embeddings')
-                    .delete()
-                    .eq('id', id)
-                    .eq('persona_id', this.personaId);
-            }
-            catch {
-                // Non-fatal: vector cleanup failure is acceptable.
-            }
-        }
-    }
-    async expireEpisodes(before, minImportance) {
-        await this.setPersonaContext();
-        const threshold = minImportance ?? 0;
-        // Count first, then delete (Supabase delete doesn't return count reliably)
-        const { count, error: countErr } = await this.client
-            .from('episodes')
-            .select('*', { count: 'exact', head: true })
-            .eq('persona_id', this.personaId)
-            .lt('timestamp', before)
-            .lt('importance', threshold);
-        if (countErr)
-            throw new Error(`expireEpisodes count failed: ${countErr.message}`);
-        const expired = count ?? 0;
-        if (expired === 0)
-            return 0;
-        const { error } = await this.client
-            .from('episodes')
-            .delete()
-            .eq('persona_id', this.personaId)
-            .lt('timestamp', before)
-            .lt('importance', threshold);
-        if (error)
-            throw new Error(`expireEpisodes failed: ${error.message}`);
-        // Best-effort: clean up orphaned embeddings
-        if (this.vectorEnabled) {
-            try {
-                await this.client.rpc('cleanup_orphaned_embeddings', {
-                    filter_persona_id: this.personaId,
-                });
-            }
-            catch {
-                // Non-fatal: orphan cleanup failure is acceptable.
-            }
-        }
-        return expired;
-    }
-    // ── Utility ──
-    async indexEmbedding(id, embedding, metadata) {
-        const { error } = await this.client.from('memory_embeddings').upsert({
-            id,
-            persona_id: this.personaId,
-            type: metadata.type,
-            source_entity: metadata.sourceEntity ?? null,
-            timestamp: metadata.timestamp,
-            importance: metadata.importance ?? 0,
-            embedding,
-        });
-        if (error)
-            throw new Error(`indexEmbedding failed: ${error.message}`);
-    }
-}
-exports.SupabaseMemoryAdapter = SupabaseMemoryAdapter;
-// ── Row mappers ──
-function rowToEpisode(row) {
-    return {
-        id: row.id,
-        timestamp: row.timestamp,
-        sourceEntity: row.source_entity || undefined,
-        context: row.context || undefined,
-        appraisal: row.appraisal,
-        emotionSnapshot: {
-            V: row.vad_v,
-            A: row.vad_a,
-            D: row.vad_d,
-        },
-        intensity: row.intensity,
-        importance: row.importance,
-    };
-}
-function rowToReflection(row) {
-    return {
-        id: row.id,
-        timestamp: row.timestamp,
-        sourceEntity: row.source_entity || undefined,
-        content: row.content,
-        trigger: row.trigger,
-        emotionSnapshot: {
-            V: row.vad_v,
-            A: row.vad_a,
-            D: row.vad_d,
-        },
-    };
-}

package/dist/cjs/memory/supabase/semantic.d.ts

@@ -1,44 +0,0 @@
-import type { EmbeddingAdapter } from '../types';
-import type { SemanticMemory, SearchOptions, SearchResult } from '../semantic';
-export interface SupabaseSemanticMemoryOptions {
-    /** Supabase project URL. */
-    supabaseUrl: string;
-    /** Supabase anon or service-role key. */
-    supabaseKey: string;
-    /** Embedding provider (embed function + dimension). */
-    embedding: EmbeddingAdapter;
-    /** Match threshold for pgvector similarity. Default: 0.7. */
-    matchThreshold?: number;
-    /** Table name for embeddings. Default: 'memory_embeddings'. */
-    tableName?: string;
-    /** RPC function name for similarity search. Default: 'match_semantic_memory'. */
-    matchFunction?: string;
-}
-/**
- * Supabase pgvector-backed SemanticMemory.
- *
- * Requires a table with columns: id (text PK), text (text), embedding (vector),
- * metadata (jsonb), and an RPC function for similarity search.
- *
- * @example
- * ```typescript
- * const memory = new SupabaseSemanticMemory({
- *   supabaseUrl: process.env.SUPABASE_URL!,
- *   supabaseKey: process.env.SUPABASE_KEY!,
- *   embedding: openaiEmbedding({ apiKey: '...' }),
- * });
- * await memory.store('user likes coffee', { userId: '123' });
- * ```
- */
-export declare class SupabaseSemanticMemory implements SemanticMemory {
-    private client;
-    private embedding;
-    private matchThreshold;
-    private tableName;
-    private matchFunction;
-    constructor(options: SupabaseSemanticMemoryOptions);
-    store(text: string, metadata?: Record<string, unknown>): Promise<string>;
-    search(query: string, options?: SearchOptions): Promise<SearchResult[]>;
-    delete(ids: string[]): Promise<void>;
-}
-//# sourceMappingURL=semantic.d.ts.map

package/dist/cjs/memory/supabase/semantic.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"semantic.d.ts","sourceRoot":"","sources":["../../../../src/memory/supabase/semantic.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,UAAU,CAAC;AACjD,OAAO,KAAK,EAAE,cAAc,EAAE,aAAa,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAE/E,MAAM,WAAW,6BAA6B;IAC5C,4BAA4B;IAC5B,WAAW,EAAE,MAAM,CAAC;IACpB,yCAAyC;IACzC,WAAW,EAAE,MAAM,CAAC;IACpB,uDAAuD;IACvD,SAAS,EAAE,gBAAgB,CAAC;IAC5B,6DAA6D;IAC7D,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,+DAA+D;IAC/D,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,iFAAiF;IACjF,aAAa,CAAC,EAAE,MAAM,CAAC;CACxB;AAED;;;;;;;;;;;;;;;GAeG;AACH,qBAAa,sBAAuB,YAAW,cAAc;IAC3D,OAAO,CAAC,MAAM,CAAiB;IAC/B,OAAO,CAAC,SAAS,CAAmB;IACpC,OAAO,CAAC,cAAc,CAAS;IAC/B,OAAO,CAAC,SAAS,CAAS;IAC1B,OAAO,CAAC,aAAa,CAAS;gBAElB,OAAO,EAAE,6BAA6B;IAQ5C,KAAK,CACT,IAAI,EAAE,MAAM,EACZ,QAAQ,GAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAM,GACrC,OAAO,CAAC,MAAM,CAAC;IAcZ,MAAM,CACV,KAAK,EAAE,MAAM,EACb,OAAO,GAAE,aAAkB,GAC1B,OAAO,CAAC,YAAY,EAAE,CAAC;IAwBpB,MAAM,CAAC,GAAG,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC;CAQ3C"}

package/dist/cjs/memory/supabase/semantic.js

@@ -1,72 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.SupabaseSemanticMemory = void 0;
-const supabase_js_1 = require("@supabase/supabase-js");
-/**
- * Supabase pgvector-backed SemanticMemory.
- *
- * Requires a table with columns: id (text PK), text (text), embedding (vector),
- * metadata (jsonb), and an RPC function for similarity search.
- *
- * @example
- * ```typescript
- * const memory = new SupabaseSemanticMemory({
- *   supabaseUrl: process.env.SUPABASE_URL!,
- *   supabaseKey: process.env.SUPABASE_KEY!,
- *   embedding: openaiEmbedding({ apiKey: '...' }),
- * });
- * await memory.store('user likes coffee', { userId: '123' });
- * ```
- */
-class SupabaseSemanticMemory {
-    constructor(options) {
-        this.client = (0, supabase_js_1.createClient)(options.supabaseUrl, options.supabaseKey);
-        this.embedding = options.embedding;
-        this.matchThreshold = options.matchThreshold ?? 0.7;
-        this.tableName = options.tableName ?? 'memory_embeddings';
-        this.matchFunction = options.matchFunction ?? 'match_semantic_memory';
-    }
-    async store(text, metadata = {}) {
-        const id = crypto.randomUUID();
-        const vector = await this.embedding.embed(text);
-        const { error } = await this.client.from(this.tableName).upsert({
-            id,
-            text,
-            embedding: vector,
-            metadata,
-        });
-        if (error)
-            throw new Error(`store failed: ${error.message}`);
-        return id;
-    }
-    async search(query, options = {}) {
-        const topK = options.topK ?? 5;
-        const threshold = options.scoreThreshold ?? this.matchThreshold;
-        const vector = await this.embedding.embed(query);
-        const { data, error } = await this.client.rpc(this.matchFunction, {
-            query_embedding: vector,
-            match_threshold: threshold,
-            match_count: topK,
-            filter_metadata: options.filter ?? null,
-        });
-        if (error)
-            throw new Error(`search failed: ${error.message}`);
-        return (data ?? []).map((row) => ({
-            id: row.id,
-            text: row.text,
-            score: row.similarity,
-            metadata: row.metadata ?? {},
-        }));
-    }
-    async delete(ids) {
-        if (ids.length === 0)
-            return;
-        const { error } = await this.client
-            .from(this.tableName)
-            .delete()
-            .in('id', ids);
-        if (error)
-            throw new Error(`delete failed: ${error.message}`);
-    }
-}
-exports.SupabaseSemanticMemory = SupabaseSemanticMemory;

package/dist/esm/embedding/cloudflare.d.ts

@@ -1,15 +0,0 @@
-import type { EmbeddingAdapter } from '../memory/types';
-export interface AiBinding {
-    run(model: string, input: {
-        text: string[];
-    }): Promise<{
-        data: number[][];
-    }>;
-}
-export interface CloudflareAIEmbeddingOptions {
-    ai: AiBinding;
-    model?: string;
-    dimension?: number;
-}
-export declare function cloudflareAIEmbedding(options: CloudflareAIEmbeddingOptions): EmbeddingAdapter;
-//# sourceMappingURL=cloudflare.d.ts.map

package/dist/esm/embedding/cloudflare.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"cloudflare.d.ts","sourceRoot":"","sources":["../../../src/embedding/cloudflare.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAExD,MAAM,WAAW,SAAS;IACxB,GAAG,CACD,KAAK,EAAE,MAAM,EACb,KAAK,EAAE;QAAE,IAAI,EAAE,MAAM,EAAE,CAAA;KAAE,GACxB,OAAO,CAAC;QAAE,IAAI,EAAE,MAAM,EAAE,EAAE,CAAA;KAAE,CAAC,CAAC;CAClC;AAED,MAAM,WAAW,4BAA4B;IAC3C,EAAE,EAAE,SAAS,CAAC;IACd,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAKD,wBAAgB,qBAAqB,CACnC,OAAO,EAAE,4BAA4B,GACpC,gBAAgB,CAWlB"}

package/dist/esm/embedding/cloudflare.js

@@ -1,13 +0,0 @@
-const DEFAULT_MODEL = '@cf/baai/bge-base-en-v1.5';
-const DEFAULT_DIMENSION = 768;
-export function cloudflareAIEmbedding(options) {
-    const model = options.model ?? DEFAULT_MODEL;
-    const dimension = options.dimension ?? DEFAULT_DIMENSION;
-    return {
-        dimension,
-        async embed(text) {
-            const result = await options.ai.run(model, { text: [text] });
-            return result.data[0];
-        },
-    };
-}

package/dist/esm/embedding/cohere.d.ts

@@ -1,8 +0,0 @@
-import type { EmbeddingAdapter } from '../memory/types';
-export interface CohereEmbeddingOptions {
-    apiKey: string;
-    model?: string;
-    inputType?: 'search_document' | 'search_query';
-}
-export declare function cohereEmbedding(options: CohereEmbeddingOptions): EmbeddingAdapter;
-//# sourceMappingURL=cohere.d.ts.map

package/dist/esm/embedding/cohere.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"cohere.d.ts","sourceRoot":"","sources":["../../../src/embedding/cohere.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAExD,MAAM,WAAW,sBAAsB;IACrC,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,iBAAiB,GAAG,cAAc,CAAC;CAChD;AAED,wBAAgB,eAAe,CAAC,OAAO,EAAE,sBAAsB,GAAG,gBAAgB,CA4BjF"}

package/dist/esm/embedding/cohere.js

@@ -1,28 +0,0 @@
-export function cohereEmbedding(options) {
-    const model = options.model ?? 'embed-english-v3.0';
-    const inputType = options.inputType ?? 'search_document';
-    const dimension = model.includes('v3') ? 1024 : 768;
-    return {
-        dimension,
-        embed: async (text) => {
-            const res = await fetch('https://api.cohere.com/v2/embed', {
-                method: 'POST',
-                headers: {
-                    'Authorization': `Bearer ${options.apiKey}`,
-                    'Content-Type': 'application/json',
-                },
-                body: JSON.stringify({
-                    model,
-                    texts: [text],
-                    input_type: inputType,
-                    embedding_types: ['float'],
-                }),
-            });
-            if (!res.ok) {
-                throw new Error(`Cohere embedding failed: ${res.status} ${res.statusText}`);
-            }
-            const json = await res.json();
-            return json.embeddings.float[0];
-        },
-    };
-}

package/dist/esm/embedding/index.d.ts

@@ -1,9 +0,0 @@
-export { openaiEmbedding } from './openai';
-export { cohereEmbedding } from './cohere';
-export { createLocalEmbedder } from './local';
-export { cloudflareAIEmbedding } from './cloudflare';
-export type { OpenAIEmbeddingOptions } from './openai';
-export type { CohereEmbeddingOptions } from './cohere';
-export type { LocalEmbedderOptions } from './local';
-export type { CloudflareAIEmbeddingOptions, AiBinding } from './cloudflare';
-//# sourceMappingURL=index.d.ts.map

package/dist/esm/embedding/index.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/embedding/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,eAAe,EAAE,MAAM,UAAU,CAAC;AAC3C,OAAO,EAAE,eAAe,EAAE,MAAM,UAAU,CAAC;AAC3C,OAAO,EAAE,mBAAmB,EAAE,MAAM,SAAS,CAAC;AAC9C,OAAO,EAAE,qBAAqB,EAAE,MAAM,cAAc,CAAC;AACrD,YAAY,EAAE,sBAAsB,EAAE,MAAM,UAAU,CAAC;AACvD,YAAY,EAAE,sBAAsB,EAAE,MAAM,UAAU,CAAC;AACvD,YAAY,EAAE,oBAAoB,EAAE,MAAM,SAAS,CAAC;AACpD,YAAY,EAAE,4BAA4B,EAAE,SAAS,EAAE,MAAM,cAAc,CAAC"}

package/dist/esm/embedding/index.js

@@ -1,4 +0,0 @@
-export { openaiEmbedding } from './openai';
-export { cohereEmbedding } from './cohere';
-export { createLocalEmbedder } from './local';
-export { cloudflareAIEmbedding } from './cloudflare';

package/dist/esm/embedding/local.d.ts

@@ -1,6 +0,0 @@
-import type { EmbeddingAdapter } from '../memory/types';
-export interface LocalEmbedderOptions {
-    model?: string;
-}
-export declare function createLocalEmbedder(options?: LocalEmbedderOptions): Promise<EmbeddingAdapter>;
-//# sourceMappingURL=local.d.ts.map

package/dist/esm/embedding/local.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"local.d.ts","sourceRoot":"","sources":["../../../src/embedding/local.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAExD,MAAM,WAAW,oBAAoB;IACnC,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;AAED,wBAAsB,mBAAmB,CACvC,OAAO,CAAC,EAAE,oBAAoB,GAC7B,OAAO,CAAC,gBAAgB,CAAC,CA4B3B"}

package/dist/esm/embedding/local.js

@@ -1,25 +0,0 @@
-export async function createLocalEmbedder(options) {
-    const modelId = options?.model ?? 'Xenova/all-MiniLM-L6-v2';
-    let pipeline;
-    try {
-        const transformers = await Function('return import("@huggingface/transformers")')();
-        pipeline = transformers.pipeline;
-    }
-    catch {
-        throw new Error('@huggingface/transformers is required for local embeddings. ' +
-            'Install it: npm install @huggingface/transformers');
-    }
-    const extractor = await pipeline('feature-extraction', modelId);
-    const probe = await extractor('test', { pooling: 'mean', normalize: true });
-    const dimension = probe.data.length;
-    return {
-        dimension,
-        embed: async (text) => {
-            const output = await extractor(text, {
-                pooling: 'mean',
-                normalize: true,
-            });
-            return Array.from(output.data);
-        },
-    };
-}

package/dist/esm/embedding/openai.d.ts

@@ -1,9 +0,0 @@
-import type { EmbeddingAdapter } from '../memory/types';
-export interface OpenAIEmbeddingOptions {
-    apiKey: string;
-    model?: string;
-    dimension?: number;
-    baseUrl?: string;
-}
-export declare function openaiEmbedding(options: OpenAIEmbeddingOptions): EmbeddingAdapter;
-//# sourceMappingURL=openai.d.ts.map

package/dist/esm/embedding/openai.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"openai.d.ts","sourceRoot":"","sources":["../../../src/embedding/openai.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAExD,MAAM,WAAW,sBAAsB;IACrC,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,wBAAgB,eAAe,CAAC,OAAO,EAAE,sBAAsB,GAAG,gBAAgB,CAuBjF"}

package/dist/esm/embedding/openai.js

@@ -1,23 +0,0 @@
-export function openaiEmbedding(options) {
-    const model = options.model ?? 'text-embedding-3-small';
-    const dimension = options.dimension ?? 1536;
-    const baseUrl = options.baseUrl ?? 'https://api.openai.com';
-    return {
-        dimension,
-        embed: async (text) => {
-            const res = await fetch(`${baseUrl}/v1/embeddings`, {
-                method: 'POST',
-                headers: {
-                    'Authorization': `Bearer ${options.apiKey}`,
-                    'Content-Type': 'application/json',
-                },
-                body: JSON.stringify({ model, input: text, dimensions: dimension }),
-            });
-            if (!res.ok) {
-                throw new Error(`OpenAI embedding failed: ${res.status} ${res.statusText}`);
-            }
-            const json = await res.json();
-            return json.data[0].embedding;
-        },
-    };
-}

package/dist/esm/events/console.d.ts

@@ -1,25 +0,0 @@
-import type { EventAdapter, SDKEvent } from './types';
-export interface ConsoleEventAdapterOptions {
-    /** Prefix for log lines. Default: '[molroo]'. */
-    prefix?: string;
-    /** Only emit these event types. Empty = all. */
-    filter?: string[];
-    /** Output format. Default: 'human'. */
-    format?: 'human' | 'json';
-    /** Custom logger function. Default: console.log. */
-    logger?: (message: string) => void;
-}
-/**
- * Console-based EventAdapter for development/debugging.
- * Logs events to console (or a custom logger function).
- */
-export declare class ConsoleEventAdapter implements EventAdapter {
-    private prefix;
-    private filter;
-    private format;
-    private logger;
-    constructor(options?: ConsoleEventAdapterOptions);
-    emit(event: SDKEvent): Promise<void>;
-    emitBatch(events: SDKEvent[]): Promise<void>;
-}
-//# sourceMappingURL=console.d.ts.map

package/dist/esm/events/console.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"console.d.ts","sourceRoot":"","sources":["../../../src/events/console.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAE,QAAQ,EAAE,MAAM,SAAS,CAAC;AAEtD,MAAM,WAAW,0BAA0B;IACzC,iDAAiD;IACjD,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,gDAAgD;IAChD,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,uCAAuC;IACvC,MAAM,CAAC,EAAE,OAAO,GAAG,MAAM,CAAC;IAC1B,oDAAoD;IACpD,MAAM,CAAC,EAAE,CAAC,OAAO,EAAE,MAAM,KAAK,IAAI,CAAC;CACpC;AAED;;;GAGG;AACH,qBAAa,mBAAoB,YAAW,YAAY;IACtD,OAAO,CAAC,MAAM,CAAS;IACvB,OAAO,CAAC,MAAM,CAAc;IAC5B,OAAO,CAAC,MAAM,CAAmB;IACjC,OAAO,CAAC,MAAM,CAA4B;gBAE9B,OAAO,CAAC,EAAE,0BAA0B;IAO1C,IAAI,CAAC,KAAK,EAAE,QAAQ,GAAG,OAAO,CAAC,IAAI,CAAC;IAapC,SAAS,CAAC,MAAM,EAAE,QAAQ,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC;CAKnD"}