anaslimem

anaslimem / CortexaDB

Public

It is a simple, fast, and hard-durable embedded database designed specifically for AI agent memory. It provides a single-file-like experience (no server required) but with native support for vectors, graphs, and temporal search.

agentic-memory ai-agents graph-database python rust
20
1
100% credibility
Found Feb 26, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

CortexaDB is a single-file embedded database designed for AI agent memory with support for vector similarity, graph relationships, temporal queries, and document ingestion.

How It Works

1
📖 Discover CortexaDB

You hear about a simple tool that gives AI helpers perfect long-term memory, like a digital notebook that never forgets.

2
⬇️ Get it ready

Download and set up the memory tool in moments, no hassle.

3
💾 Save memories

Tell your AI facts like 'User loves dark mode' or feed whole documents, and it stores them safely.

4
🗣️ Ask questions

Simply ask 'What does the user like?' and get the best matching memories instantly.

5
🔗 Connect ideas

Link related facts together so your AI understands relationships and context.

🎉 AI remembers perfectly

Your AI helper now recalls everything from chats to files, making it smarter every time.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 20 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 CortexaDB?

CortexaDB is a simple, fast embedded database built in Rust for storing AI agent memory in a single file—no server needed. It handles text memories with automatic vector embeddings, graph connections between them, and temporal queries by recency, plus smart chunking for loading docs like PDFs or Markdown. Python users get a dead-simple API: open a .mem file, call remember("user likes dark mode"), load files, ask semantic questions, and connect related memories.

Why is it gaining traction?

Unlike bloated vector DBs requiring clusters, it's a drop-in single file with hybrid search blending semantics, graphs, and time—perfect for agents needing quick recall without setup hassle. Hard durability via WAL survives crashes, auto-evicts old entries, and supports replay for debugging, all in a simple fast package that feels like SQLite for embeddings. Python bindings and PyPI install make it agent-ready out of the box.

Who should use this?

AI devs building autonomous agents that need persistent, searchable memory, like chatbots remembering user prefs or multi-step task planners linking facts. Solo hackers prototyping LLM sidekicks or researchers testing agent memory without infra overhead. Skip if you need massive scale—it's for embedded, local agent brains.

Verdict

Promising beta for simple agent memory (11 stars, 1.0% credibility), with solid Python docs and tests, but low adoption means watch for stability. Try it if you're tired of juggling separate vector/graph stores—pip install and prototype in minutes.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.