v3-memory-unification

Name: v3-memory-unification
Author: spencermarx

301stars24forks

How to install v3-memory-unification?

spencermarx/open-code-review/v3-memory-unification

$npx -y skills add spencermarx/open-code-review --skill v3-memory-unification

Installs into the current project.

Summary

Unify 6+ memory systems into AgentDB with HNSW indexing for 150x-12,500x search improvements. Implements ADR-006 (Unified Memory Service) and ADR-009 (Hybrid Memory Backend).

Files · 1

View on GitHub

SKILL.md

1	# V3 Memory Unification
2
3	## What This Skill Does
4
5	Consolidates disparate memory systems into unified AgentDB backend with HNSW vector search, achieving 150x-12,500x search performance improvements while maintaining backward compatibility.
6
7	## Quick Start
8
9	```bash
10	# Initialize memory unification
11	Task("Memory architecture", "Design AgentDB unification strategy", "v3-memory-specialist")
12
13	# AgentDB integration
14	Task("AgentDB setup", "Configure HNSW indexing and vector search", "v3-memory-specialist")
15
16	# Data migration
17	Task("Memory migration", "Migrate SQLite/Markdown to AgentDB", "v3-memory-specialist")
18	```
19
20	## Systems to Unify
21
22	### Legacy Systems → AgentDB
23	```
24	┌─────────────────────────────────────────┐
25	│ • MemoryManager (basic operations) │
26	│ • DistributedMemorySystem (clustering) │
27	│ • SwarmMemory (agent-specific) │
28	│ • AdvancedMemoryManager (features) │
29	│ • SQLiteBackend (structured) │
30	│ • MarkdownBackend (file-based) │
31	│ • HybridBackend (combination) │
32	└─────────────────────────────────────────┘
33	↓
34	┌─────────────────────────────────────────┐
35	│ 🚀 AgentDB with HNSW │
36	│ • 150x-12,500x faster search │
37	│ • Unified query interface │
38	│ • Cross-agent memory sharing │
39	│ • SONA learning integration │
40	└─────────────────────────────────────────┘
41	```
42
43	## Implementation Architecture
44
45	### Unified Memory Service
46	```typescript
47	class UnifiedMemoryService implements IMemoryBackend {
48	constructor(
49	private agentdb: AgentDBAdapter,
50	private indexer: HNSWIndexer,
51	private migrator: DataMigrator
52	) {}
53
54	async store(entry: MemoryEntry): Promise<void> {
55	await this.agentdb.store(entry);
56	await this.indexer.index(entry);
57	}
58
59	async query(query: MemoryQuery): Promise<MemoryEntry[]> {
60	if (query.semantic) {
61	return this.indexer.search(query); // 150x-12,500x faster
62	}
63	return this.agentdb.query(query);
64	}
65	}
66	```
67
68	### HNSW Vector Search
69	```typescript
70	class HNSWIndexer {
71	constructor(dimensions: number = 1536) {
72	this.index = new HNSWIndex({
73	dimensions,
74	efConstruction: 200,
75	M: 16,
76	speedupTarget: '150x-12500x'
77	});
78	}
79
80	async search(query: MemoryQuery): Promise<MemoryEntry[]> {
81	const embedding = await this.embedContent(query.content);
82	const results = this.index.search(embedding, query.limit \|\| 10);
83	return this.retrieveEntries(results);
84	}
85	}
86	```
87
88	## Migration Strategy
89
90	### Phase 1: Foundation
91	```typescript
92	// AgentDB adapter setup
93	const agentdb = new AgentDBAdapter({
94	dimensions: 1536,
95	indexType: 'HNSW',
96	speedupTarget: '150x-12500x'
97	});
98	```
99
100	### Phase 2: Data Migration
101	```typescript
102	// SQLite → AgentDB
103	const migrateFromSQLite = async () => {
104	const entries = await sqlite.getAll();
105	for (const entry of entries) {
106	const embedding = await generateEmbedding(entry.content);
107	await agentdb.store({ ...entry, embedding });
108	}
109	};
110
111	// Markdown → AgentDB
112	const migrateFromMarkdown = async () => {
113	const files = await glob('*/.md');
114	for (const file of files) {
115	const content = await fs.readFile(file, 'utf-8');
116	await agentdb.store({
117	id: generateId(),
118	content,
119	embedding: await generateEmbedding(content),
120	metadata: { originalFile: file }
121	});
122	}
123	};
124	```
125
126	## SONA Integration
127
128	### Learning Pattern Storage
129	```typescript
130	class SONAMemoryIntegration {
131	async storePattern(pattern: LearningPattern): Promise<void> {
132	await this.memory.store({
133	id: pattern.id,
134	content: pattern.data,
135	metadata: {
136	sonaMode: pattern.mode,
137	reward: pattern.reward,
138	adaptationTime: pattern.adaptationTime
139	},
140	embedding: await this.generateEmbedding(pattern.data)
141	});
142	}
143
144	async retrieveSimilarPatterns(query: string): Promise<LearningPattern[]> {
145	return this.memory.query({
146	type: 'semantic',
147	content: query,
148	filters: { type: 'learning_pattern' }
149	});
150	}
151	}
152	```
153
154	## Performance Targets
155
156	- Search Speed: 150x-12,500x improvement via HNSW
157	- Memory Usage: 50-75% reduction through optimization
158	- Query Latency: <100ms for 1M+ entries
159	- Cross-Agent Sharing: Real-time memory synchronization
160	- SONA Integration: <0.05ms adaptation time
161
162	## Success Metrics
163
164	- [ ] All 7 legacy memory systems migrated to AgentDB
165	- [ ] 150x-12,500x search performance validated
166	- [ ] 50-75% memory usage reduction achieved
167	- [ ] Backward compatibility maintained
168	- [ ] SONA learning patterns integrated
169	- [ ] Cross-agent memory sharing operational

People who liked this also viewed

6 picks

Type

v3-memory-unification

301stars24forks

View on GitHub

How to install v3-memory-unification?

spencermarx/open-code-review/v3-memory-unification

$npx -y skills add spencermarx/open-code-review --skill v3-memory-unification

Installs into the current project.

Summary

Unify 6+ memory systems into AgentDB with HNSW indexing for 150x-12,500x search improvements. Implements ADR-006 (Unified Memory Service) and ADR-009 (Hybrid Memory Backend).

Files · 1

View on GitHub

SKILL.md

1	# V3 Memory Unification
2
3	## What This Skill Does
4
5	Consolidates disparate memory systems into unified AgentDB backend with HNSW vector search, achieving 150x-12,500x search performance improvements while maintaining backward compatibility.
6
7	## Quick Start
8
9	```bash
10	# Initialize memory unification
11	Task("Memory architecture", "Design AgentDB unification strategy", "v3-memory-specialist")
12
13	# AgentDB integration
14	Task("AgentDB setup", "Configure HNSW indexing and vector search", "v3-memory-specialist")
15
16	# Data migration
17	Task("Memory migration", "Migrate SQLite/Markdown to AgentDB", "v3-memory-specialist")
18	```
19
20	## Systems to Unify
21
22	### Legacy Systems → AgentDB
23	```
24	┌─────────────────────────────────────────┐
25	│ • MemoryManager (basic operations) │
26	│ • DistributedMemorySystem (clustering) │
27	│ • SwarmMemory (agent-specific) │
28	│ • AdvancedMemoryManager (features) │
29	│ • SQLiteBackend (structured) │
30	│ • MarkdownBackend (file-based) │
31	│ • HybridBackend (combination) │
32	└─────────────────────────────────────────┘
33	↓
34	┌─────────────────────────────────────────┐
35	│ 🚀 AgentDB with HNSW │
36	│ • 150x-12,500x faster search │
37	│ • Unified query interface │
38	│ • Cross-agent memory sharing │
39	│ • SONA learning integration │
40	└─────────────────────────────────────────┘
41	```
42
43	## Implementation Architecture
44
45	### Unified Memory Service
46	```typescript
47	class UnifiedMemoryService implements IMemoryBackend {
48	constructor(
49	private agentdb: AgentDBAdapter,
50	private indexer: HNSWIndexer,
51	private migrator: DataMigrator
52	) {}
53
54	async store(entry: MemoryEntry): Promise<void> {
55	await this.agentdb.store(entry);
56	await this.indexer.index(entry);
57	}
58
59	async query(query: MemoryQuery): Promise<MemoryEntry[]> {
60	if (query.semantic) {
61	return this.indexer.search(query); // 150x-12,500x faster
62	}
63	return this.agentdb.query(query);
64	}
65	}
66	```
67
68	### HNSW Vector Search
69	```typescript
70	class HNSWIndexer {
71	constructor(dimensions: number = 1536) {
72	this.index = new HNSWIndex({
73	dimensions,
74	efConstruction: 200,
75	M: 16,
76	speedupTarget: '150x-12500x'
77	});
78	}
79
80	async search(query: MemoryQuery): Promise<MemoryEntry[]> {
81	const embedding = await this.embedContent(query.content);
82	const results = this.index.search(embedding, query.limit \|\| 10);
83	return this.retrieveEntries(results);
84	}
85	}
86	```
87
88	## Migration Strategy
89
90	### Phase 1: Foundation
91	```typescript
92	// AgentDB adapter setup
93	const agentdb = new AgentDBAdapter({
94	dimensions: 1536,
95	indexType: 'HNSW',
96	speedupTarget: '150x-12500x'
97	});
98	```
99
100	### Phase 2: Data Migration
101	```typescript
102	// SQLite → AgentDB
103	const migrateFromSQLite = async () => {
104	const entries = await sqlite.getAll();
105	for (const entry of entries) {
106	const embedding = await generateEmbedding(entry.content);
107	await agentdb.store({ ...entry, embedding });
108	}
109	};
110
111	// Markdown → AgentDB
112	const migrateFromMarkdown = async () => {
113	const files = await glob('*/.md');
114	for (const file of files) {
115	const content = await fs.readFile(file, 'utf-8');
116	await agentdb.store({
117	id: generateId(),
118	content,
119	embedding: await generateEmbedding(content),
120	metadata: { originalFile: file }
121	});
122	}
123	};
124	```
125
126	## SONA Integration
127
128	### Learning Pattern Storage
129	```typescript
130	class SONAMemoryIntegration {
131	async storePattern(pattern: LearningPattern): Promise<void> {
132	await this.memory.store({
133	id: pattern.id,
134	content: pattern.data,
135	metadata: {
136	sonaMode: pattern.mode,
137	reward: pattern.reward,
138	adaptationTime: pattern.adaptationTime
139	},
140	embedding: await this.generateEmbedding(pattern.data)
141	});
142	}
143
144	async retrieveSimilarPatterns(query: string): Promise<LearningPattern[]> {
145	return this.memory.query({
146	type: 'semantic',
147	content: query,
148	filters: { type: 'learning_pattern' }
149	});
150	}
151	}
152	```
153
154	## Performance Targets
155
156	- Search Speed: 150x-12,500x improvement via HNSW
157	- Memory Usage: 50-75% reduction through optimization
158	- Query Latency: <100ms for 1M+ entries
159	- Cross-Agent Sharing: Real-time memory synchronization
160	- SONA Integration: <0.05ms adaptation time
161
162	## Success Metrics
163
164	- [ ] All 7 legacy memory systems migrated to AgentDB
165	- [ ] 150x-12,500x search performance validated
166	- [ ] 50-75% memory usage reduction achieved
167	- [ ] Backward compatibility maintained
168	- [ ] SONA learning patterns integrated
169	- [ ] Cross-agent memory sharing operational

People who liked this also viewed

6 picks

Type