# v3-memory-specialist

The v3-memory-specialist subagent for Claude Code migrates scattered project memory into AgentDB with HNSW indexing, speeding semantic search and sharing context across agents.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 850 repo stars
- Author: ruvnet
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/ruvnet/agentic-flow/tree/HEAD/.claude/agents/v3/v3-memory-specialist.md
- Updated: 2026-10-26
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/v3-memory-specialist

## What it does

The agent coordinates migrating seven disparate memory components, such as SQLite, markdown storage, and swarm memory, into a single `UnifiedMemoryService`. It swaps linear O(n) memory scans for logarithmic HNSW nearest-neighbor indexing.

The architecture handles hybrid retrieval: structured queries filter entries by agent metadata, while semantic searches match embeddings. Pre- and post-execution hooks depend on `agentic-flow@alpha` via `npx`, so tracking patterns will fail without a working Node.js runtime.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/v3-memory-specialist.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/v3-memory-specialist.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/v3-memory-specialist.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/ruvnet/agentic-flow/tree/HEAD/.claude/agents/v3/v3-memory-specialist.md

## When to use it
- migrate memory entries from SQLite and Markdown backends into AgentDB
- replace linear lookups with HNSW vector indexing for faster semantic search
- enable shared context across multiple autonomous agents
- record learning patterns and evaluation trajectories in the SONA layer

## Example prompts
- Plan the migration of existing SQLite memory records into AgentDB using HNSW indexing
- Configure semantic memory search for agent context with a threshold of 0.8
- Store the agent learning pattern and execution trajectory in unified memory

## Similar
- [memory-specialist](https://affpapa.org/en/nevibecode/agents/memory-specialist.md) — subagent · ★ 189 repo stars
- [V3 Memory Unification](https://affpapa.org/en/nevibecode/skills/v3-memory-unification.md) — skill · 1100 installs
- [AgentDB Memory Patterns](https://affpapa.org/en/nevibecode/skills/agentdb-memory-patterns.md) — skill · 1100 installs
- [agentdb-persistent-memory-patterns](https://affpapa.org/en/nevibecode/skills/agentdb-persistent-memory-patterns.md) — skill · 1 install
- [Holographic Memory](https://affpapa.org/en/nevibecode/mcp/holographic-memory.md) — MCP server · ★ 23 repo stars
- [vector-index-tuning](https://affpapa.org/en/nevibecode/skills/vector-index-tuning.md) — skill · 10700 installs

From the same set: [ruvnet/agentic-flow](https://affpapa.org/en/nevibecode/by/ruvnet/agentic-flow.md) — 76 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
