AVIDS2

AVIDS2 / memorix

Public

Cross-Agent Memory Bridge Persistent memory for AI coding agents across Cursor, Windsurf, Claude Code, Codex, Copilot, Kiro via MCP. Never re-explain your project again.

162
12
100% credibility
Found Feb 24, 2026 at 55 stars 3x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Memorix provides persistent shared memory for AI coding assistants across multiple tools, enabling recall of project decisions, fixes, and patterns without re-explaining.

How It Works

1
😩 Tired of repeating yourself

Every new chat with your AI coding helper, you explain the project again from scratch.

2
🔍 Discover Memorix

You find Memorix, a helper that makes your AI remember everything across chats and different tools.

3
Connect with one setting

Add a single line to your AI tool's preferences, and restart—it just works without extra accounts or costs.

4
🧠 AI starts remembering

Now your helper recalls past decisions, fixes, and tips automatically, saving you time every session.

5
📱 View your memory dashboard

Open a simple web page to see your project's knowledge map, recent notes, and smart cleanup suggestions.

6
🔄 Switch tools seamlessly

Move your setup to another AI coding app with one command—everything syncs over instantly.

🎉 Never forget again

Your AI feels like a teammate that knows your project deeply, across any tool or session.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 55 to 162 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 memorix?

Memorix is a TypeScript-based MCP server that acts as a cross-agent memory bridge for AI coding agents like Cursor, Windsurf, Claude Code, Codex, Copilot, and Kiro. It stores project decisions, gotchas, and fixes persistently on disk, so agents recall your architecture across sessions without re-explaining again. Add it to your MCP config with one npx command, and tools like memorix_search or memorix_store handle retrieval via progressive disclosure.

Why is it gaining traction?

Unlike single-agent memory tools, memorix syncs workspaces—MCP configs, rules, and skills—across agents, plus auto-hooks capture insights implicitly. The interactive dashboard visualizes knowledge graphs with D3.js, and CLI commands like memorix_workspace_sync migrate setups in seconds. Developers notice seamless context when switching from Cursor to Claude Code, with local vector search optional for semantic matches.

Who should use this?

AI-heavy coders juggling Cursor and Copilot for daily tasks, or backend teams using Windsurf for dev and Claude Code for reviews. Ideal for solo devs tired of pasting context every session, or onboarding new members where agents instantly know gotchas like Windows path bugs.

Verdict

Try memorix if you run multiple agents—its 422 passing tests and full docs make setup reliable despite 27 stars and 1.0% credibility score. Still early; watch for polish on Antigravity quirks, but Apache 2.0 invites forks.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.