git404hub

what is mneme-open fr?

axisrobo/mneme-open — explained in plain English

Analysis updated 2026-05-18

11PythonAudience · developerComplexity · 3/5LicenseSetup · moderate

tl;dr

Open source client SDKs and docs for Mneme, a memory and context system that gives AI agents long term, searchable memory.

vibe map

mindmap
  root((Mneme open))
    What it does
      Agent long term memory
      Context assembly
      Hybrid search
    Tech stack
      Python
      TypeScript
      Go
    Use cases
      Persistent agent memory
      Multi language clients
      Prebuilt server binaries
    Audience
      AI developers
      Agent builders

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

what do people make with this?

VIBE 1

Give an AI agent long term memory that survives beyond a single conversation's context window.

VIBE 2

Store agent observations, facts, and decisions as searchable, timestamped records.

VIBE 3

Connect a Python, TypeScript, or Go application to a running Mneme memory server.

VIBE 4

Prototype an agent that retrieves relevant past context using hybrid keyword and semantic search.

what's the stack?

PythonTypeScriptGogRPCJSON-RPC

how it stacks up fr

axisrobo/mneme-open2arons/llm-cliabe238/claude-video-plus
Stars111111
LanguagePythonPythonPython
Setup difficultymoderateeasyeasy
Complexity3/52/53/5
Audiencedeveloperdeveloperdeveloper

Figures from each repo's GitHub metadata at analysis time.

how do i run it?

Difficulty · moderate time til it works · 30min

The Mneme engine server itself is closed source, only a prebuilt binary or the public client SDKs are available here.

Free to use, modify, and distribute, including commercially, as long as the license notice is kept.

in plain english

Mneme-open is the open source portion of a project called Mneme, which is described as a memory system for AI agents. The idea is that an AI agent working on its own, separate from the language model it uses for thinking, needs somewhere to store long term memory, assemble relevant context, and represent knowledge in a structured way, rather than relying only on whatever fits inside a single conversation. Mneme stores every observation, fact, decision, and plan as its own typed record with a timestamp, and lets you search across all of that using a mix of keyword search, meaning based search, relationships between records, and time. This particular repository does not contain the Mneme server itself, which is closed source, but everything needed to talk to a running Mneme server: client libraries for Python, TypeScript, and Go, a command line tool, runnable example programs, full API documentation, and the schema definitions for the different ways of talking to the server, including JSON-RPC, a protocol called MCP, REST, and gRPC. Prebuilt server binaries are also published from this repository's release page, so you do not need the private engine source code to actually run a Mneme server yourself. To get started, you download a prebuilt server binary for your platform from the releases page and start it, which by default keeps everything in memory unless you point it at a location on disk to persist data instead. From there you pick whichever client language matches your project, install it, and connect to the running server's local address. The documentation includes a table comparing the three protocols by what they are best suited for: for example the JSON-RPC over HTTP option exposes the widest set of operations while gRPC offers a smaller, more strictly typed set. Everything in this repository is released under the Apache 2.0 license, while the prebuilt server binaries carry a separate license of their own.

prompts (copy fr)

prompt 1
Help me start a local Mneme server and connect to it with the Python HTTP client.
prompt 2
Explain the difference between an episode, a fact, and a piece of knowledge in Mneme's data model.
prompt 3
Show me how to add an episode and then search for it using the Go client.
prompt 4
Compare the JSON-RPC, gRPC, and REST options in Mneme and help me pick one for my project.

Frequently asked questions

what is mneme-open fr?

Open source client SDKs and docs for Mneme, a memory and context system that gives AI agents long term, searchable memory.

What language is mneme-open written in?

Mainly Python. The stack also includes Python, TypeScript, Go.

What license does mneme-open use?

Free to use, modify, and distribute, including commercially, as long as the license notice is kept.

How hard is mneme-open to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is mneme-open for?

Mainly developer.

peek the repo → explain another one

This repo across BitVibe Labs

double-check against the repo, no cap.