skills.rest
33may
33mayCOMMUNITY
Explore all skills by this author →
GitHub

V3 Memory Unification

Consolidates multiple memory backends into AgentDB with HNSW vector search indexing.

1|1|Updated Nov 28, 2025
Search Tags:#semantic search#embeddings#agentdb#vector search#hnsw#unification#memory migration
One-click install
npx skills add https://github.com/33may/robotics --skill v3-memory-unification-33may
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: V3 Memory Unification
Source: https://github.com/33may/robotics/tree/main/humanoid/.claude/skills/v3-memory-unification
Command: npx skills add https://github.com/33may/robotics --skill v3-memory-unification-33may

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Fragmented memory systems (SQLite, Markdown, swarm memory, distributed stores) create inconsistent query interfaces and slow retrieval. This Skill unifies them into a single AgentDB backend with HNSW vector indexing for semantic search. ## Core Features & Use Cases - Unified Memory Service: Implements a single IMemoryBackend interface routing semantic queries to HNSW search and structured queries to AgentDB. - Data Migration: Migrates entries from SQLite backends and Markdown files into AgentDB with generated embeddings. - SONA Integration: Stores and retrieves learning patterns with reward and adaptation metadata for cross-agent memory sharing. - Use Case: An agent system running seven separate memory stores can migrate all entries into AgentDB, then run semantic queries across 1M+ entries with sub-100ms latency. ## Quick Start Ask the agent to design an AgentDB unification strategy and migrate the existing SQLite and Markdown memory stores into a single HNSW-indexed backend.

Frequently Asked Questions about V3 Memory Unification

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I migrate SQLite memory data to AgentDB?▼

Read all entries from the SQLite backend, generate an embedding for each entry's content, then store the entry with its embedding into AgentDB. The migration runs as a loop over existing records, preserving metadata during transfer.

What is HNSW indexing used for in memory search?▼

HNSW (Hierarchical Navigable Small World) indexing enables fast approximate nearest-neighbor search over vector embeddings. It is configured with parameters like dimensions, efConstruction, and M to deliver large search speedups over brute-force queries.

Can Markdown files be used as a memory backend?▼

Yes, Markdown files can be migrated into AgentDB by reading each file, generating an embedding from its content, and storing it with metadata referencing the original file path. After migration, semantic search works across all file content.

Does the unified memory service keep backward compatibility?▼

Yes, the UnifiedMemoryService implements the existing IMemoryBackend interface, so legacy callers continue working. Semantic queries route to the HNSW indexer while standard queries go directly to AgentDB.

What are the limitations of vector-based memory search?▼

Vector search depends on embedding quality and index configuration, so poorly embedded content may return weak matches. Structured exact-match queries should still route to the standard AgentDB query path rather than semantic search.

Related Skills

Explore More Software Engineering Skills→
fabri07fabri07

V3 Memory Unification

Consolidate SQLite, Markdown, and distributed memory stores into AgentDB with HNSW vector search.

Community
Advanced
dzhechkodzhechko

V3 Memory Unification

Unify multiple memory backends into AgentDB with HNSW vector search.

Community
Advanced
novatech2210-cmdnovatech2210-cmd

V3 Performance Optimization

Benchmark and optimize Claude v3 performance with Flash Attention and HNSW indexing.

Community
Advanced
mahg-esmahg-es
3

vector-search

Configure HNSW/IVF indices and perform ANN search with vector embeddings.

Official
Advanced