Skip to main content

Universal memory provider for AI agents (Redis + Kumiho)

Project description

Kumiho Memory


Experimental client-side utilities for AI agent memory integration


⚠️ Status

Experimental / Preview (0.1.x) This package is provided for early experimentation and reference usage. APIs and behavior may change without notice. Latest patch: 0.1.2 (2026-02-09) - README refresh and version metadata sync.


What this package is

kumiho-memory provides client-side utilities that help AI agents temporarily buffer interaction context and interface with the broader Kumiho Cognitive Memory architecture.

It is designed to be:

  • Lightweight
  • Model-agnostic
  • Framework-agnostic
  • Safe to use in local or sandboxed environments

What this package is NOT

To avoid confusion, this package does NOT implement:

  • ❌ A full cognitive memory system
  • ❌ Long-term memory graphs or lineage tracking
  • ❌ Memory consolidation or offline processing
  • ❌ Automated belief revision or pruning
  • ❌ The "Dream State" consolidation pipeline

Those capabilities exist at the system level and are intentionally decoupled from this client-side library.


Design intent

This separation is intentional.

By keeping advanced memory logic outside the client library:

  • Memory remains independent of any specific LLM
  • Client environments stay fast and lightweight
  • Sensitive or irreversible memory operations are centrally controlled
  • The architecture remains portable across platforms and models

Typical use cases

  • Experimenting with memory-aware AI agents

  • Prototyping agent workflows that require short-term context buffering

  • Reference integration for platforms such as:

    • Multi-agent systems
    • Collaborative AI environments
    • MCP-compatible agent runtimes

Installation

pip install kumiho-memory

Minimal example

from kumiho_memory import RedisMemoryBuffer

memory = RedisMemoryBuffer()

memory.add_message(
    project="example",
    session_id="demo-session",
    role="user",
    content="Hello!"
)

This example demonstrates temporary, short-term buffering only. It does not represent long-term memory persistence.


Architectural note

kumiho-memory is one component within a larger, model-agnostic memory architecture.

The full system includes:

  • Client-side buffers (this package)
  • Persistent memory storage
  • Structured relationships between memories
  • Offline consolidation and lifecycle management

This package intentionally exposes only the client-side surface.


Roadmap

  • 0.1.x — Experimental preview (current)
  • 0.2.x — Stabilized client APIs
  • 1.0.0 — Production-ready client SDK

The scope of this package will remain limited to client-side concerns.


License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kumiho_memory-0.3.0.tar.gz (66.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

kumiho_memory-0.3.0-py3-none-any.whl (54.4 kB view details)

Uploaded Python 3

File details

Details for the file kumiho_memory-0.3.0.tar.gz.

File metadata

  • Download URL: kumiho_memory-0.3.0.tar.gz
  • Upload date:
  • Size: 66.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.14

File hashes

Hashes for kumiho_memory-0.3.0.tar.gz
Algorithm Hash digest
SHA256 f07b9b6caa8087034e4c5819a04f91278e4f475c6ed449d897fc0dc28a7624e1
MD5 74d5cced2e3c7695f0996ec34fac5b3b
BLAKE2b-256 6bd51518f8c4567553ca3ad735d6e947f1c2a4a619676b668a5e06df22453ea5

See more details on using hashes here.

File details

Details for the file kumiho_memory-0.3.0-py3-none-any.whl.

File metadata

  • Download URL: kumiho_memory-0.3.0-py3-none-any.whl
  • Upload date:
  • Size: 54.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.14

File hashes

Hashes for kumiho_memory-0.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 5b799b24e6f3f1c14c7e523e3f106dee52ec62b91a5bcbb2ac880ce4657dca4f
MD5 c134f4e8e62fd2cf347e15d660f30c45
BLAKE2b-256 50964631a98a0d715418fdcb9dc4126dcf394641f2dea24c9f54376800e28bb7

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page