> ## Documentation Index
> Fetch the complete documentation index at: https://docs.infino.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent memory

> Give AI agents long-term memory with Infino. Hybrid BM25 and vector recall over past conversations, ready-made as the memory-infino OpenClaw plugin.

[`memory-infino`](https://clawhub.ai/infino-ai/plugins/memory-infino) is a drop-in
memory plugin for [OpenClaw](https://docs.openclaw.ai) agents, built on Infino. An agent
stores what it sees and recalls the relevant pieces later, searched by keyword and by
meaning together. It's all queryable with SQL and kept on object storage.

## The memory-infino plugin

[`memory-infino`](https://clawhub.ai/infino-ai/plugins/memory-infino) implements OpenClaw's
memory-plugin contract with three tools:

| Tool            | What it does                                                                           |
| --------------- | -------------------------------------------------------------------------------------- |
| `memory_recall` | Hybrid (BM25 + vector) search over long-term memory, fused with reciprocal-rank fusion |
| `memory_store`  | Save a memory (embedded on write)                                                      |
| `memory_forget` | Delete by id, or find-and-delete by query                                              |

## Install

```bash theme={null}
npm install @infino-ai/memory-infino
```

Find it on [ClawHub](https://clawhub.ai/infino-ai/plugins/memory-infino) and enable it as
a memory plugin in your OpenClaw configuration. See the
[OpenClaw docs](https://docs.openclaw.ai) for wiring a memory plugin into an agent. Memory
lives on a local path or object storage (S3 / GCS / Azure).

## Why Infino for agent memory

* **Hybrid recall in one engine.** `memory_recall` fuses keyword (BM25) and semantic
  (vector) matches, so an agent finds memories by exact terms and by meaning at once, with no
  separate keyword index or rerank service.
* **Object-storage-native.** Memory is stored as Apache Parquet on a local path or object
  storage; it scales with the agent's history rather than a fixed cluster.
* **SQL over memory.** The same store answers structured and time-based questions over an
  agent's history.

## Build it yourself

Not on OpenClaw, or want memory tailored to your own stack? Build the same pattern
directly on the Infino Node binding. See the
[agent-memory tutorial](/tutorials) and the [Quickstart](/quickstart).

## See also

* [memory-infino on ClawHub](https://clawhub.ai/infino-ai/plugins/memory-infino)
* [memory-infino on GitHub](https://github.com/infino-ai/memory-infino)
* [How Infino works](/core-concepts)