@soulcraft/brainy 0.63.0 → 1.0.0-rc.2

package/dist/chat/brainyChat.d.ts

@@ -1,42 +0,0 @@
-/**
- * Brainy Chat - Talk to Your Data
- *
- * Simple, powerful conversational AI for your Brainy database.
- * Works with zero configuration, optionally enhanced with LLM.
- */
-import { BrainyData } from '../brainyData.js';
-export interface ChatOptions {
-    /** Optional LLM model name or provider:model format */
-    llm?: string;
-    /** Include source references in responses */
-    sources?: boolean;
-    /** API key for LLM provider (if needed) */
-    apiKey?: string;
-}
-export declare class BrainyChat {
-    private brainy;
-    private llmProvider?;
-    private options;
-    private history;
-    constructor(brainy: BrainyData, options?: ChatOptions);
-    /**
-     * Initialize LLM provider based on model string
-     */
-    private initializeLLM;
-    /**
-     * Ask a question - works with or without LLM
-     */
-    ask(question: string): Promise<string>;
-    /**
-     * Generate response using LLM
-     */
-    private generateWithLLM;
-    /**
-     * Generate response with smart templates (no LLM needed)
-     */
-    private generateWithTemplate;
-    /**
-     * Interactive chat mode (Node.js only)
-     */
-    chat(): Promise<void>;
-}

package/dist/chat/brainyChat.js

@@ -1,340 +0,0 @@
-/**
- * Brainy Chat - Talk to Your Data
- *
- * Simple, powerful conversational AI for your Brainy database.
- * Works with zero configuration, optionally enhanced with LLM.
- */
-export class BrainyChat {
-    constructor(brainy, options = {}) {
-        this.history = [];
-        this.brainy = brainy;
-        this.options = options;
-        // Load LLM if specified
-        if (options.llm) {
-            this.initializeLLM(options.llm, options.apiKey);
-        }
-    }
-    /**
-     * Initialize LLM provider based on model string
-     */
-    async initializeLLM(model, apiKey) {
-        // Parse provider from model string (e.g., "claude-3-5-sonnet", "gpt-4", "Xenova/LaMini")
-        if (model.startsWith('claude') || model.includes('anthropic')) {
-            this.llmProvider = new ClaudeLLMProvider(model, apiKey);
-        }
-        else if (model.startsWith('gpt') || model.includes('openai')) {
-            this.llmProvider = new OpenAILLMProvider(model, apiKey);
-        }
-        else if (model.includes('/')) {
-            // Hugging Face model format
-            this.llmProvider = new HuggingFaceLLMProvider(model);
-        }
-        else {
-            console.warn(`Unknown LLM model: ${model}, falling back to templates`);
-        }
-    }
-    /**
-     * Ask a question - works with or without LLM
-     */
-    async ask(question) {
-        // Find relevant context using vector search
-        const searchResults = await this.brainy.search(question, 10);
-        // Generate response
-        let answer;
-        if (this.llmProvider) {
-            answer = await this.generateWithLLM(question, searchResults);
-        }
-        else {
-            answer = this.generateWithTemplate(question, searchResults);
-        }
-        // Add sources if requested
-        if (this.options.sources && searchResults.length > 0) {
-            const sources = searchResults
-                .slice(0, 3)
-                .map(r => r.id)
-                .join(', ');
-            answer += `\n[Sources: ${sources}]`;
-        }
-        // Track history (keep last 10 exchanges)
-        this.history.push({ question, answer });
-        if (this.history.length > 10) {
-            this.history = this.history.slice(-10);
-        }
-        return answer;
-    }
-    /**
-     * Generate response using LLM
-     */
-    async generateWithLLM(question, context) {
-        if (!this.llmProvider) {
-            return this.generateWithTemplate(question, context);
-        }
-        // Build context from search results
-        const contextData = context.map(item => ({
-            id: item.id,
-            score: item.score,
-            metadata: item.metadata || {}
-        }));
-        // Include conversation history for context
-        const historyContext = this.history.slice(-3).map(h => `Q: ${h.question}\nA: ${h.answer}`).join('\n\n');
-        try {
-            const response = await this.llmProvider.generate(question, {
-                searchResults: contextData,
-                history: historyContext
-            });
-            return response;
-        }
-        catch (error) {
-            console.warn('LLM generation failed, using template:', error);
-            return this.generateWithTemplate(question, context);
-        }
-    }
-    /**
-     * Generate response with smart templates (no LLM needed)
-     */
-    generateWithTemplate(question, context) {
-        if (context.length === 0) {
-            return "I couldn't find relevant information to answer that question.";
-        }
-        const q = question.toLowerCase();
-        // Quantitative questions
-        if (q.includes('how many') || q.includes('count')) {
-            const count = context.length;
-            const items = context.slice(0, 3).map(c => c.id).join(', ');
-            return `I found ${count} relevant items. The top matches are: ${items}.`;
-        }
-        // Comparison questions
-        if (q.includes('compare') || q.includes('difference') || q.includes('vs')) {
-            if (context.length < 2) {
-                return "I need at least two items to make a comparison.";
-            }
-            const first = context[0];
-            const second = context[1];
-            return `Comparing "${first.id}" (${(first.score * 100).toFixed(0)}% relevance) with "${second.id}" (${(second.score * 100).toFixed(0)}% relevance). Both are related to your query but ${first.id} shows stronger similarity.`;
-        }
-        // List questions
-        if (q.includes('list') || q.includes('what are') || q.includes('show me')) {
-            const items = context.slice(0, 5).map((c, i) => `${i + 1}. ${c.id}${c.metadata?.description ? ': ' + c.metadata.description : ''}`).join('\n');
-            return `Here are the top results:\n${items}`;
-        }
-        // Analysis questions
-        if (q.includes('analyze') || q.includes('explain') || q.includes('why')) {
-            const top = context[0];
-            const metadata = top.metadata || {};
-            const details = Object.entries(metadata)
-                .slice(0, 3)
-                .map(([k, v]) => `${k}: ${JSON.stringify(v)}`)
-                .join(', ');
-            return `Based on my analysis of "${top.id}" (${(top.score * 100).toFixed(0)}% relevant): ${details || 'This item matches your query based on semantic similarity.'}`;
-        }
-        // Trend/pattern questions
-        if (q.includes('trend') || q.includes('pattern')) {
-            const items = context.slice(0, 3).map(c => c.id);
-            return `I identified patterns across ${context.length} related items. Key examples include: ${items.join(', ')}. These show common characteristics related to "${question}".`;
-        }
-        // Yes/No questions
-        if (q.startsWith('is') || q.startsWith('are') || q.startsWith('does') || q.startsWith('do')) {
-            const confidence = context[0].score;
-            if (confidence > 0.8) {
-                return `Yes, based on "${context[0].id}" with ${(confidence * 100).toFixed(0)}% confidence.`;
-            }
-            else if (confidence > 0.5) {
-                return `Possibly. I found "${context[0].id}" with ${(confidence * 100).toFixed(0)}% relevance to your question.`;
-            }
-            else {
-                return `I'm not certain. The closest match is "${context[0].id}" but with only ${(confidence * 100).toFixed(0)}% relevance.`;
-            }
-        }
-        // Default response - provide the most relevant information
-        const top = context[0];
-        const metadata = top.metadata ?
-            Object.entries(top.metadata)
-                .slice(0, 3)
-                .map(([k, v]) => `${k}: ${JSON.stringify(v)}`)
-                .join(', ') :
-            'no additional details';
-        return `Based on "${top.id}" (${(top.score * 100).toFixed(0)}% relevant): ${metadata}`;
-    }
-    /**
-     * Interactive chat mode (Node.js only)
-     */
-    async chat() {
-        // Check if we're in Node.js
-        if (typeof process === 'undefined' || !process.stdin) {
-            console.log('Interactive chat is only available in Node.js environment');
-            return;
-        }
-        const readline = await import('readline');
-        const rl = readline.createInterface({
-            input: process.stdin,
-            output: process.stdout,
-            prompt: 'You> '
-        });
-        console.log('\n🧠 Brainy Chat - Interactive Mode');
-        console.log('Type your questions or "exit" to quit\n');
-        rl.prompt();
-        rl.on('line', async (line) => {
-            const input = line.trim();
-            if (input.toLowerCase() === 'exit' || input.toLowerCase() === 'quit') {
-                console.log('\nGoodbye! 👋');
-                rl.close();
-                return;
-            }
-            if (input) {
-                try {
-                    const answer = await this.ask(input);
-                    console.log(`\n🤖 ${answer}\n`);
-                }
-                catch (error) {
-                    console.log(`\n❌ Error: ${error instanceof Error ? error.message : String(error)}\n`);
-                }
-            }
-            rl.prompt();
-        });
-        rl.on('close', () => {
-            process.exit(0);
-        });
-    }
-}
-/**
- * Claude LLM Provider
- */
-class ClaudeLLMProvider {
-    constructor(model, apiKey) {
-        this.model = model.includes('claude') ? model : `claude-3-5-sonnet-20241022`;
-        this.apiKey = apiKey || process.env.ANTHROPIC_API_KEY;
-    }
-    async generate(prompt, context) {
-        if (!this.apiKey) {
-            throw new Error('Claude API key required. Set ANTHROPIC_API_KEY or pass apiKey option.');
-        }
-        const systemPrompt = `You are a helpful AI assistant with access to a vector database. 
-Answer questions based on the provided context from semantic search results.
-Be concise and accurate. If the context doesn't contain relevant information, say so.`;
-        const userPrompt = `Context from database search:
-${JSON.stringify(context.searchResults, null, 2)}
-
-Recent conversation:
-${context.history || 'No previous conversation'}
-
-Question: ${prompt}
-
-Please provide a helpful answer based on the context above.`;
-        try {
-            const response = await fetch('https://api.anthropic.com/v1/messages', {
-                method: 'POST',
-                headers: {
-                    'Content-Type': 'application/json',
-                    'x-api-key': this.apiKey,
-                    'anthropic-version': '2023-06-01'
-                },
-                body: JSON.stringify({
-                    model: this.model,
-                    max_tokens: 1024,
-                    messages: [
-                        { role: 'user', content: userPrompt }
-                    ],
-                    system: systemPrompt
-                })
-            });
-            if (!response.ok) {
-                throw new Error(`Claude API error: ${response.status}`);
-            }
-            const data = await response.json();
-            return data.content[0].text;
-        }
-        catch (error) {
-            throw new Error(`Failed to generate with Claude: ${error instanceof Error ? error.message : String(error)}`);
-        }
-    }
-}
-/**
- * OpenAI LLM Provider
- */
-class OpenAILLMProvider {
-    constructor(model, apiKey) {
-        this.model = model.includes('gpt') ? model : 'gpt-4o-mini';
-        this.apiKey = apiKey || process.env.OPENAI_API_KEY;
-    }
-    async generate(prompt, context) {
-        if (!this.apiKey) {
-            throw new Error('OpenAI API key required. Set OPENAI_API_KEY or pass apiKey option.');
-        }
-        const systemPrompt = `You are a helpful AI assistant with access to a vector database. 
-Answer questions based on the provided context from semantic search results.`;
-        const userPrompt = `Context: ${JSON.stringify(context.searchResults)}
-History: ${context.history || 'None'}
-Question: ${prompt}`;
-        try {
-            const response = await fetch('https://api.openai.com/v1/chat/completions', {
-                method: 'POST',
-                headers: {
-                    'Content-Type': 'application/json',
-                    'Authorization': `Bearer ${this.apiKey}`
-                },
-                body: JSON.stringify({
-                    model: this.model,
-                    messages: [
-                        { role: 'system', content: systemPrompt },
-                        { role: 'user', content: userPrompt }
-                    ],
-                    max_tokens: 500,
-                    temperature: 0.7
-                })
-            });
-            if (!response.ok) {
-                throw new Error(`OpenAI API error: ${response.status}`);
-            }
-            const data = await response.json();
-            return data.choices[0].message.content;
-        }
-        catch (error) {
-            throw new Error(`Failed to generate with OpenAI: ${error instanceof Error ? error.message : String(error)}`);
-        }
-    }
-}
-/**
- * Hugging Face Local LLM Provider
- */
-class HuggingFaceLLMProvider {
-    constructor(model) {
-        this.model = model;
-        this.initializePipeline();
-    }
-    async initializePipeline() {
-        try {
-            // Lazy load transformers.js - this is optional and may not be installed
-            // @ts-ignore - Optional dependency
-            const transformersModule = await import('@huggingface/transformers').catch(() => null);
-            if (transformersModule) {
-                const { pipeline } = transformersModule;
-                this.pipeline = await pipeline('text2text-generation', this.model);
-            }
-            else {
-                console.warn(`Transformers.js not installed. Install with: npm install @huggingface/transformers`);
-            }
-        }
-        catch (error) {
-            console.warn(`Failed to load Hugging Face model ${this.model}:`, error);
-        }
-    }
-    async generate(prompt, context) {
-        if (!this.pipeline) {
-            throw new Error('Hugging Face model not loaded');
-        }
-        const input = `Answer based on context: ${JSON.stringify(context.searchResults).slice(0, 500)}
-Question: ${prompt}
-Answer:`;
-        try {
-            const result = await this.pipeline(input, {
-                max_new_tokens: 150,
-                temperature: 0.7
-            });
-            return result[0].generated_text.trim();
-        }
-        catch (error) {
-            throw new Error(`Failed to generate with Hugging Face: ${error instanceof Error ? error.message : String(error)}`);
-        }
-    }
-}
-//# sourceMappingURL=brainyChat.js.map

package/dist/cortex/cliWrapper.d.ts

@@ -1,32 +0,0 @@
-/**
- * CLI Wrapper for Cortex
- *
- * Provides CLI-specific functionality that wraps the core Cortex orchestrator
- * Following our philosophy: "Simple for beginners, powerful for experts"
- */
-export declare class CortexCLI {
-    private brainy;
-    private cortex;
-    private neuralImport?;
-    constructor();
-    /**
-     * Initialize the brain and cortex
-     */
-    init(options?: any): Promise<boolean>;
-    /**
-     * Add data - with Neural Import processing by default
-     */
-    add(data?: string, metadata?: any): Promise<void>;
-    /**
-     * Search with helpful display
-     */
-    search(query?: string, options?: any): Promise<void>;
-    /**
-     * Show statistics
-     */
-    stats(detailed?: boolean): Promise<void>;
-    /**
-     * Interactive chat interface
-     */
-    chat(initialMessage?: string): Promise<void>;
-}

package/dist/cortex/cliWrapper.js

@@ -1,209 +0,0 @@
-/**
- * CLI Wrapper for Cortex
- *
- * Provides CLI-specific functionality that wraps the core Cortex orchestrator
- * Following our philosophy: "Simple for beginners, powerful for experts"
- */
-import { BrainyData } from '../brainyData.js';
-import { Cortex } from '../cortex.js';
-import { NeuralImport } from './neuralImport.js';
-// @ts-ignore
-import chalk from 'chalk';
-// @ts-ignore
-import ora from 'ora';
-// @ts-ignore
-import prompts from 'prompts';
-export class CortexCLI {
-    constructor() {
-        this.brainy = new BrainyData();
-        this.cortex = new Cortex();
-    }
-    /**
-     * Initialize the brain and cortex
-     */
-    async init(options = {}) {
-        const spinner = ora('Initializing brain...').start();
-        try {
-            // Initialize BrainyData
-            await this.brainy.init();
-            // Set up Neural Import as default augmentation
-            this.neuralImport = new NeuralImport(this.brainy);
-            this.cortex.register(this.neuralImport);
-            spinner.succeed('Brain initialized!');
-            return true;
-        }
-        catch (error) {
-            spinner.fail('Initialization failed');
-            throw error;
-        }
-    }
-    /**
-     * Add data - with Neural Import processing by default
-     */
-    async add(data, metadata) {
-        // If no data provided, go interactive
-        if (!data) {
-            const input = await prompts([
-                {
-                    type: 'text',
-                    name: 'data',
-                    message: 'What would you like to add?',
-                    validate: (value) => value.length > 0 || 'Please enter something'
-                }
-            ]);
-            if (!input.data)
-                return;
-            data = input.data;
-        }
-        // Check if it's a file path
-        const fs = await import('fs');
-        if (fs.existsSync(data)) {
-            // Use Neural Import for files
-            if (this.neuralImport) {
-                console.log(chalk.cyan('🧠 Using Neural Import to understand your data...'));
-                const result = await this.neuralImport.neuralImport(data, {
-                    autoApply: true,
-                    previewOnly: false
-                });
-                console.log(chalk.green(`✅ Imported ${result.detectedEntities.length} entities`));
-                console.log(chalk.gray(`   ${result.detectedRelationships.length} relationships detected`));
-                console.log(chalk.gray(`   Confidence: ${(result.confidence * 100).toFixed(1)}%`));
-                if (result.insights.length > 0) {
-                    console.log(chalk.cyan('\n💡 Insights:'));
-                    result.insights.forEach(insight => {
-                        console.log(chalk.gray(`   • ${insight.description}`));
-                    });
-                }
-            }
-        }
-        else {
-            // Plain text - add directly
-            const id = await this.brainy.add(data, metadata);
-            console.log(chalk.green('✅ Added!'));
-            console.log(chalk.gray(`   ID: ${id}`));
-        }
-    }
-    /**
-     * Search with helpful display
-     */
-    async search(query, options = {}) {
-        // Interactive mode if no query
-        if (!query) {
-            const input = await prompts({
-                type: 'text',
-                name: 'query',
-                message: 'What are you looking for?',
-                validate: (value) => value.length > 0 || 'Please enter a search term'
-            });
-            if (!input.query)
-                return;
-            query = input.query;
-        }
-        const results = await this.brainy.search(query, options.limit || 5);
-        if (results.length === 0) {
-            console.log(chalk.yellow('No results found'));
-            console.log(chalk.gray('Try different keywords or add more data'));
-        }
-        else {
-            console.log(chalk.cyan(`\nFound ${results.length} results:\n`));
-            results.forEach((result, i) => {
-                const content = result.content || result.text || result.data;
-                console.log(chalk.white(`${i + 1}. ${content}`));
-                console.log(chalk.gray(`   Score: ${(result.score * 100).toFixed(1)}%`));
-                if (result.metadata) {
-                    const keys = Object.keys(result.metadata).slice(0, 3);
-                    if (keys.length > 0) {
-                        const preview = keys.map(k => `${k}: ${result.metadata[k]}`).join(', ');
-                        console.log(chalk.gray(`   ${preview}`));
-                    }
-                }
-                console.log();
-            });
-        }
-    }
-    /**
-     * Show statistics
-     */
-    async stats(detailed = false) {
-        const stats = await this.brainy.getStats();
-        console.log(chalk.cyan('\n🧠 Brain Statistics\n'));
-        console.log(chalk.white('Capacity:'));
-        console.log(`  Memories: ${stats.totalNouns || 0}`);
-        console.log(`  Connections: ${stats.totalVerbs || 0}`);
-        if (detailed) {
-            console.log(chalk.white('\nBreakdown:'));
-            if (stats.nounTypes) {
-                console.log('  Entity Types:');
-                Object.entries(stats.nounTypes).forEach(([type, count]) => {
-                    console.log(`    ${type}: ${count}`);
-                });
-            }
-            if (stats.verbTypes) {
-                console.log('  Relationship Types:');
-                Object.entries(stats.verbTypes).forEach(([type, count]) => {
-                    console.log(`    ${type}: ${count}`);
-                });
-            }
-        }
-        console.log(chalk.white('\nAugmentations:'));
-        const augmentations = this.cortex.getAugmentationsByType('sense');
-        console.log(`  Active: ${augmentations.length}`);
-        if (augmentations.length > 0) {
-            augmentations.forEach(aug => {
-                console.log(`    • ${aug.name || 'unnamed'}`);
-            });
-        }
-    }
-    /**
-     * Interactive chat interface
-     */
-    async chat(initialMessage) {
-        console.log(chalk.cyan('\n🧠 Brain Chat'));
-        console.log(chalk.gray('I can help you explore your data. Type "exit" to quit.\n'));
-        // Note: Full chat would integrate with BrainyChat
-        // For now, we use search as a simple implementation
-        const readline = (await import('readline')).createInterface({
-            input: process.stdin,
-            output: process.stdout,
-            prompt: chalk.cyan('You: ')
-        });
-        if (initialMessage) {
-            console.log(chalk.cyan('You: ') + initialMessage);
-            const results = await this.brainy.search(initialMessage, 3);
-            if (results.length > 0) {
-                console.log(chalk.green('Brain: ') + 'Based on what I know:');
-                results.forEach(r => {
-                    console.log(chalk.gray(`  • ${r.content || r.text}`));
-                });
-            }
-            else {
-                console.log(chalk.green('Brain: ') + "I don't have information about that yet.");
-            }
-            console.log();
-        }
-        readline.prompt();
-        readline.on('line', async (line) => {
-            const input = line.trim();
-            if (input.toLowerCase() === 'exit') {
-                console.log(chalk.gray('\nGoodbye! 👋'));
-                readline.close();
-                process.exit(0);
-            }
-            if (input) {
-                const results = await this.brainy.search(input, 3);
-                if (results.length > 0) {
-                    console.log(chalk.green('Brain: ') + 'I found:');
-                    results.forEach(r => {
-                        console.log(chalk.gray(`  • ${r.content || r.text}`));
-                    });
-                }
-                else {
-                    console.log(chalk.green('Brain: ') + "I don't know about that yet.");
-                }
-                console.log();
-            }
-            readline.prompt();
-        });
-    }
-}
-//# sourceMappingURL=cliWrapper.js.map