dx111ge

dx111ge / engram

Public

AI Memory Engine -- knowledge graph + semantic search + reasoning + learning in a single binary

ai-memory graph-database inference-engine knowledge-graph llm-tools
11
1
100% credibility
Found Mar 11, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

Engram is a standalone tool that builds a visual knowledge graph in one file, allowing users to store facts, search semantically, reason logically, and interact via a web dashboard.

How It Works

1
🔍 Discover Engram

You find this cool memory tool online and download the simple program to your computer.

2
🧠 Make your memory file

You create a single file that acts like your personal brain to store all your ideas and facts.

3
🚀 Open the explorer

Start the program and visit the web page it opens to see your dashboard ready to go.

4
🌐 Build your knowledge web

Add people, places, or ideas, connect them like puzzle pieces, and watch the colorful graph come alive.

5
🔎 Search and chat naturally

Type everyday questions or words, and get instant answers pulled from your growing knowledge.

6
📚 Teach and refine

Tell it new facts, boost sure ones, fix errors, and let memories strengthen or fade naturally.

🎉 Smart companion ready

Your personal knowledge helper now remembers, reasons, and grows smarter with every use.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 11 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 engram?

Engram is a single-binary AI memory engine that packs a knowledge graph, hybrid semantic search (BM25 + vectors), logical reasoning, and confidence-based learning into one portable .brain file—no databases, Docker, or cloud needed. Run CLI commands like `engram store "PostgreSQL"` or `engram search "database confidence>0.5"` to build and query your graph, or fire up `engram serve` for a web UI with graph explorer and natural language queries. Built in JavaScript, it supports HTTP/gRPC APIs, MCP for Claude/Cursor, and offline ONNX embeddings, solving the hassle of scattered AI memory tools like github memory allocators or managers.

Why is it gaining traction?

It stands out with zero deps and a single-file knowledge base—copy your .brain for instant backup or migration—plus peer-to-peer sync via ed25519 trust scoring. Devs dig the built-in web dashboard for visualizing graphs, reinforcing facts, or running inference rules, without spinning up extra services. Hybrid search and lifecycle features (decay, correct, derive) make it a bright memory engine for evolving AI knowledge, unlike rigid graph DBs or basic github memory optimizers.

Who should use this?

AI engineers prototyping RAG pipelines or local LLMs needing persistent recall without vendor lock-in. OSINT analysts building threat intel graphs from Wikipedia/docs, or support teams mapping errors-to-solutions. Claude/Cursor powerusers wanting MCP tools for `engram_prove` and natural language "tell" commands to teach custom engrams.

Verdict

Promising for dep-free AI memory like a memory engine ai with ClickHouse vibes, but at 10 stars and 1.0% credibility, it's early-stage—docs shine with use cases, yet expect bugs in edge cases. Try for side projects if single-binary hooks you; skip for production until more battle-tested.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.