LLM-VLM-GSL

LLM-VLM-GSL / AriadneMem

Public

Code for Paper: AriadneMem: Threading the Maze of Lifelong Memory for LLM Agents

226
33
100% credibility
Found Feb 09, 2026 at 27 stars 8x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

AriadneMem is a graph-based memory system designed to help AI agents retain and reason over long-term conversational history with accurate state updates and multi-hop connections.

How It Works

1
🔍 Discover AriadneMem

You hear about a smart memory helper that lets your AI chat buddy remember long talks perfectly, even when plans change.

2
📥 Grab it easily

Download the handy tool with a simple command, like picking up a new app.

3
🔗 Link your AI friend

Tell it which smart AI brain to use, like connecting to your favorite thinking service.

4
💬 Share your chats

Feed in real conversations, like 'meet at 2pm' then 'change to 3pm', and watch it build a memory web.

5
🧠 Weave the memories

Click to organize everything into a clever web that tracks changes and connections.

6
❓ Ask tricky questions

Pose questions like 'What time are we meeting?' and get spot-on answers that connect the dots.

✅ AI remembers forever

Your AI now recalls details across chats flawlessly, making every conversation smarter and seamless.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 27 to 226 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is AriadneMem?

AriadneMem is a Python-based memory system for LLM agents that builds a lifelong graph from dialogue streams, threading disconnected facts and state updates into queryable paths. Feed it conversations via simple API calls like add_dialogues() and ask(), and it handles multi-hop reasoning over long histories without forgetting changes like "meet at 2pm" updated to "3pm". It runs on CPU/GPU with OpenAI/Qwen APIs and local embeddings, plus code github python repository integration via MCP servers for Cursor or HTTP clients.

Why is it gaining traction?

Unlike flat RAG or multi-round planning that racks up LLM calls and loses context, AriadneMem uses a two-phase pipeline for stable retrieval—one LLM call per query with algorithmic bridges and paths—cutting latency while boosting accuracy on benchmarks like LoCoMo. Devs dig the plug-and-play MCP setup for code github copilot, Claude, or Cursor chats, plus eco/pro modes to tune token costs. Demos and quick_test.py show multi-hop wins instantly, making it a smart pick for agents over code paper please baselines.

Who should use this?

AI devs building persistent agents for chat apps or code github ai workflows, especially those hitting memory walls in long sessions. Cursor users wanting MCP tools for project history recall, or agent builders testing LoCoMo-style tasks like causal chains in conversations. Skip if you're doing one-shot queries—it's for lifelong state tracking.

Verdict

Grab it for agent prototypes if you're in code github repository experiments; solid README, tests, and MCP make setup fast despite 38 stars and 1.0% credibility score signaling early maturity. Under active dev for code/math domains, so watch for v1 stability.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.