#!/usr/bin/env tsx
/**
* Comprehensive diagnostic for all reported issues
*/
import { initDatabaseFromEnv } from './src/database/index.js';
import { MemoryCore } from './src/core/index.js';
async function diagnoseAllIssues() {
console.log('š Comprehensive Issue Diagnosis\n');
const db = initDatabaseFromEnv();
const memoryCore = new MemoryCore(db, process.env.OPENAI_API_KEY);
await memoryCore.initialize();
const userId = 'bob@matsuoka.com';
// 1. Test Metadata Field Updates
console.log('1ļøā£ METADATA FIELD UPDATES:');
// Create a memory with initial metadata
const createResult = await memoryCore.addMemory(
'Metadata Update Test',
'Testing metadata field updates',
'semantic' as any,
{
userId,
metadata: {
project: 'alpha',
status: 'active',
version: '1.0'
},
generateEmbedding: false,
}
);
if (createResult.status === 'success' && createResult.data) {
const memId = (createResult.data as any).id;
console.log(` ā
Created memory with metadata: ${memId}`);
// Try to update metadata
const updateResult = await memoryCore.updateMemory(
memId,
{
metadata: {
project: 'alpha',
status: 'completed', // Changed
version: '1.1', // Changed
newField: 'test' // Added
}
} as any,
{ userId }
);
console.log(` Update result: ${updateResult.status}`);
if (updateResult.status === 'error') {
console.log(` ā Metadata update not supported: ${updateResult.error}`);
}
// Check what fields can be updated
const getResult = await memoryCore.getMemory(memId, { userId });
if (getResult.status === 'success' && getResult.data) {
const memory = getResult.data as any;
console.log(' Current metadata:', JSON.stringify(memory.metadata));
}
// Clean up
await memoryCore.deleteMemory(memId, { userId });
}
// 2. Test Metadata-Based Search
console.log('\n2ļøā£ METADATA-BASED SEARCH:');
// Create test memories
const mem1 = await memoryCore.addMemory(
'Project Alpha Doc',
'Documentation for Alpha',
'semantic' as any,
{
userId,
metadata: { project: 'alpha', type: 'doc' },
generateEmbedding: false,
}
);
const mem2 = await memoryCore.addMemory(
'Project Beta Doc',
'Documentation for Beta',
'semantic' as any,
{
userId,
metadata: { project: 'beta', type: 'doc' },
generateEmbedding: false,
}
);
const memIds = [];
if (mem1.status === 'success') memIds.push((mem1.data as any).id);
if (mem2.status === 'success') memIds.push((mem2.data as any).id);
// Test metadata search
const searchTests = [
'project:alpha',
'type:doc',
'metadata.project:beta'
];
for (const query of searchTests) {
const searchResult = await memoryCore.searchMemories(query, { userId });
console.log(` Query "${query}": ${searchResult.status}`);
if (searchResult.status === 'success' && searchResult.data) {
const results = searchResult.data as any[];
console.log(` Found ${results.length} results`);
if (results.length > 0) {
console.log(` First result: ${results[0].title}`);
}
} else if (searchResult.status === 'error') {
console.log(` Error: ${searchResult.error}`);
}
}
// Clean up
for (const id of memIds) {
await memoryCore.deleteMemory(id, { userId });
}
// 3. Test Semantic Similarity Search
console.log('\n3ļøā£ SEMANTIC SIMILARITY SEARCH:');
// Create memories for similarity search
const sim1 = await memoryCore.addMemory(
'Machine Learning Basics',
'Introduction to neural networks and deep learning concepts',
'semantic' as any,
{
userId,
generateEmbedding: true, // Request embedding
}
);
const sim2 = await memoryCore.addMemory(
'AI and Neural Networks',
'Deep learning with neural network architectures',
'semantic' as any,
{
userId,
generateEmbedding: true, // Request embedding
}
);
const sim3 = await memoryCore.addMemory(
'Coffee Brewing Guide',
'How to make the perfect cup of coffee',
'semantic' as any,
{
userId,
generateEmbedding: true, // Request embedding
}
);
const simIds = [];
if (sim1.status === 'success') simIds.push((sim1.data as any).id);
if (sim2.status === 'success') simIds.push((sim2.data as any).id);
if (sim3.status === 'success') simIds.push((sim3.data as any).id);
// Search for similar content
const similaritySearch = await memoryCore.searchMemories(
'artificial intelligence and machine learning',
{
userId,
strategy: 'similarity',
threshold: 0.5,
}
);
console.log(` Similarity search status: ${similaritySearch.status}`);
if (similaritySearch.status === 'success' && similaritySearch.data) {
const results = similaritySearch.data as any[];
console.log(` Found ${results.length} similar memories`);
results.forEach((r, i) => {
console.log(` ${i + 1}. ${r.title}`);
});
}
// Clean up
for (const id of simIds) {
await memoryCore.deleteMemory(id, { userId });
}
// 4. Test Vector Embedding Generation
console.log('\n4ļøā£ VECTOR EMBEDDING GENERATION:');
console.log(` OpenAI API Key: ${process.env.OPENAI_API_KEY ? 'ā
Present' : 'ā Missing'}`);
if (process.env.OPENAI_API_KEY) {
const embedTest = await memoryCore.addMemory(
'Embedding Test Memory',
'This should generate an embedding vector',
'semantic' as any,
{
userId,
generateEmbedding: true,
}
);
if (embedTest.status === 'success' && embedTest.data) {
const embedId = (embedTest.data as any).id;
// Check database directly for embedding
const embedCheck = await db.execute(
`SELECT
id,
embedding IS NOT NULL as has_embedding,
LENGTH(embedding) as embed_length,
json_array_length(embedding) as array_length
FROM memories
WHERE id = ?`,
[embedId]
);
const row = embedCheck.rows[0] as any;
console.log(` Memory created: ${embedId}`);
console.log(` Has embedding: ${row?.has_embedding ? 'ā
' : 'ā'}`);
if (row?.has_embedding) {
console.log(` Embedding size: ${row.embed_length} bytes`);
console.log(` Vector dimensions: ${row.array_length || 'N/A'}`);
}
// Test if vector search works
const vectorSearch = await memoryCore.searchMemories(
'test embedding',
{ userId, strategy: 'similarity' }
);
console.log(` Vector search works: ${vectorSearch.status === 'success' ? 'ā
' : 'ā'}`);
// Clean up
await memoryCore.deleteMemory(embedId, { userId });
}
} else {
console.log(' ā ļø Cannot test - no API key configured');
}
// 5. Check updateMemory method signature
console.log('\n5ļøā£ UPDATE METHOD CAPABILITIES:');
console.log(' Checking updateMemory method...');
// Check the actual method implementation
const testMem = await memoryCore.addMemory(
'Update Test',
'Testing update capabilities',
'semantic' as any,
{ userId, generateEmbedding: false }
);
if (testMem.status === 'success' && testMem.data) {
const testId = (testMem.data as any).id;
// Test what can be updated
const updates = [
{ field: 'title', value: 'New Title' },
{ field: 'content', value: 'New Content' },
{ field: 'importance', value: 0.9 },
{ field: 'tags', value: ['test', 'update'] },
{ field: 'metadata', value: { key: 'value' } }
];
for (const update of updates) {
const updateObj = { [update.field]: update.value };
const result = await memoryCore.updateMemory(testId, updateObj as any, { userId });
console.log(` Update ${update.field}: ${result.status === 'success' ? 'ā
Supported' : 'ā Not supported'}`);
}
await memoryCore.deleteMemory(testId, { userId });
}
console.log('\nš DIAGNOSIS SUMMARY:');
console.log(' Issues found that need fixing:');
console.log(' - Metadata field updates not in updateMemory signature');
console.log(' - Metadata search may have SQL syntax issues');
console.log(' - Vector embeddings depend on OpenAI API key');
console.log(' - Similarity search requires valid embeddings');
process.exit(0);
}
diagnoseAllIssues().catch(console.error);
