Lyzr-Cognis

Lyzr-Cognis / cognis

Public

Lightweight open source version of cognis

18
0
100% credibility
Found Mar 25, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A lightweight Python library for giving AI agents local, persistent memory by extracting and searching facts from user conversations.

How It Works

1
🔍 Discover Cognis

You hear about a handy tool that helps AI chatbots remember personal details from conversations, just like a good friend would.

2
📦 Add it easily

You grab the tool with a simple download, ready to use in your project.

3
🔗 Connect smart helpers

You link up thinking services so the tool can understand and store facts.

4
🧠 Start your memory space

You create a personal memory bank tied to your name or ID, all local on your computer.

5
💬 Share chat messages

You feed in conversation turns, and it quietly extracts key facts like your job or hobbies.

6
🔎 Find what you need

You ask questions, and it pulls up the right memories super fast with smart matching.

🎉 Smart remembering AI

Your chatbot now recalls your details across chats, making every talk feel personal and helpful.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 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 cognis?

Cognis is a Python library for adding lightweight, local-first memory to AI agents. Feed it conversation messages via a simple API, and it auto-extracts facts into 13 categories like identity, work, or preferences using LLMs, then stores them with hybrid vector-plus-keyword search. Retrieve relevant context strings for your LLM prompts with zero servers—just pip install and set Gemini/OpenAI keys.

Why is it gaining traction?

It ditches heavy infra for in-process SQLite and local vectors, clocking ~500ms searches with only three tiny deps (~156MB total). Session scoping keeps short-term chats separate from global user facts, mirroring hosted platforms like Cognism without login hassles or preise. Devs love the quickstart: add messages, search queries, get context—no Docker, pure Python simplicity akin to a lightweight github client for agent memory.

Who should use this?

Python devs building conversational AI agents that need to remember user details, like personalized tutors recalling career goals or wellness habits. Backend teams prototyping RAG pipelines for customer support bots. Indie hackers wanting cognisant, local memory without cloud vector stores or cognism erfahrungen complexity.

Verdict

Solid docs and API make it dev-ready for prototypes, but 18 stars and 1.0% credibility scream early beta—test coverage exists, yet scale unproven. Use for lightweight agent experiments today; watch for production polish.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.