What does this plugin do?

It lets your agent store notes and find them again later using meaning-based search, backed by RuVector memory with HNSW search, AgentDB, and semantic retrieval.

Does search require matching the exact words I stored?

No. Retrieval is semantic, meaning the agent finds notes by what they mean, not by exact wording.

Where are memories kept?

In AgentDB, which the plugin uses alongside RuVector for storage and search.

ruflo-rag-memory

PackSearch

ruflo-rag-memory is a plugin that gives an AI agent a persistent memory backed by RuVector. The agent can store notes and find them again later using meaning-based search, so retrieval works even when the wording differs from the original note. It combines HNSW search, AgentDB, and semantic retrieval.

Unavailable. Delivery for this kind is on the roadmap — not serving yet.

Have an AI agent setup that supports plugins.

What your AI can do with it

Store notes that persist between agent sessions
Retrieve stored notes by meaning rather than exact wording
Search memory using HNSW vector search
Keep memories in AgentDB
Use RuVector memory for semantic retrieval

Getting started

Have an AI agent setup that supports plugins.
Add the ruflo-rag-memory plugin to it.
Enable the plugin so the agent can use RuVector memory.
Ask the agent to store a note, then search for it later by meaning to confirm retrieval works.

Signals

GitHub stars: 73k
Forks: 9k
Last commit: Sep 2026

Questions

What does this plugin do?: It lets your agent store notes and find them again later using meaning-based search, backed by RuVector memory with HNSW search, AgentDB, and semantic retrieval.
Does search require matching the exact words I stored?: No. Retrieval is semantic, meaning the agent finds notes by what they mean, not by exact wording.
Where are memories kept?: In AgentDB, which the plugin uses alongside RuVector for storage and search.

Advanced

Item type: plugin
Key: ruvnet-ruflo-ruflo-rag-memory
Source: github.com/ruvnet/ruflo

github.com/ruvnet/ruflo