hieuchaydi

hieuchaydi / MemoryFeed

Public

Local-first social memory search engine with browser capture, hybrid AI retrieval, and optional C++ acceleration.

docs-research-pv3h.vercel.app llm ollama transformers
10
0
100% credibility
Found May 07, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

MemoryFeed captures social media content you dwell on via a browser extension and provides local search, smart feeds, and resurfacing through a web dashboard and command line tools.

How It Works

1
🔍 Discover MemoryFeed

You hear about a handy tool that saves social media posts you spend time reading, so you can find them later without scrolling forever.

2
🚀 Quick setup

With one easy command, you launch your personal memory saver on your computer.

3
🧩 Add browser helper

You slip a tiny companion into your browser to quietly notice posts you linger on.

4
📱 Read and save

As you browse Facebook, Twitter, or TikTok and pause on fun posts or videos, they get tucked away safely.

5
💻 Open your dashboard

You visit the simple screen to see, search, or get smart suggestions from your saved moments.

6
⭐ Find old favorites

Type a fuzzy description like 'that angry cat meme' and instantly rediscover buried treasures.

🎉 Your memory boost

Now your social scrolls turn into a private treasure chest of resurfaced ideas, right when you need them.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 10 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 MemoryFeed?

MemoryFeed is a local-first social memory search engine built in Python with a React frontend. A browser extension for Chrome, Edge, Brave, and Firefox captures posts, images, and videos from platforms like Twitter, Facebook, YouTube, TikTok, and LinkedIn after a 3-second dwell time, storing everything in local SQLite and LanceDB. It delivers hybrid retrieval via full-text and semantic search, plus an Active Feed that ranks memories by personal heat, recency, and context for proactive resurfacing.

Why is it gaining traction?

Its optional C++ acceleration speeds up text normalization and ranking without compromising the Python core, making searches feel instant even on large personal archives. The CLI supports quick queries like `memoryfeed search "that angry cat meme"` or `memoryfeed feed --mode focus`, while MCP tools enable Claude or Cursor agents to query your social memory directly. No cloud dependency means true local-first control, with graceful degradation if AI providers like Gemini or Groq are offline.

Who should use this?

Social-heavy developers chasing tech insights on Twitter or LinkedIn who lose track of useful posts. Indie hackers building personal knowledge graphs without vendor lock-in. AI tinkerers integrating browser-captured context into agents via MCP for ambient retrieval.

Verdict

Try it if you want a lightweight social second brain—quickstart scripts get you running in 60 seconds, docs are solid. At 10 stars stars stars stars stars (10 stars) and 1.0% credibility, it's early alpha: expect rough edges, but the hybrid engine and acceleration hook make it worth forking for custom needs.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.