ereztdev

ereztdev / langmem-ts

Public

TypeScript first class memory primitive for AI agents. Embed → store → retrieve → inject, backed by Postgres + pgvector. Swappable embedders, consumer owns the connection, zero framework coupling. MIT.

agent agents ai embeddings langchain
11
0
100% credibility
Found Apr 23, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

langmem-ts is a TypeScript library offering memory tools for AI agents to extract facts from conversations, convert them to searchable forms, store them, and retrieve relevant ones later.

How It Works

1
👀 Discover langmem-ts

You hear about a handy tool that lets AI assistants remember important facts from conversations forever.

2
📦 Add the memory kit

You easily add this memory tool to your AI project setup.

3
📓 Set up memory notebook

You prepare a safe digital notebook where facts will be stored securely.

4
🔗 Link smart thinker

You connect a clever service that understands and processes chat words.

5
✨ Capture key facts

Chats magically turn into short, useful facts worth remembering.

6
💾 Save facts away

Facts get tucked into the notebook, ready for anytime recall.

7
🔍 Find memories instantly

Ask naturally about the past, and get the most relevant facts right away.

🎉 Smarter AI companion

Your AI now has a reliable memory, making every chat feel personal and aware.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 11 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 langmem-ts?

langmem-ts is a TypeScript-first memory primitive for AI agents, handling extract, embed, store, and retrieve workflows backed by Postgres and pgvector. It lets you distill facts from conversation turns, turn them into vectors with swappable embedders, persist them in your own Postgres setup, and query semantically for agent recall. Developers get a lightweight npm package with zero framework ties—you own connections and configs.

Why is it gaining traction?

Unlike Python-only LangMem or VC-backed Mem0, it's pure TypeScript, MIT-licensed, and agnostic to agent frameworks, making it a drop-in for Node apps or GitHub Actions workflows. The interface-driven design means easy swaps for embedders or backends, plus a reference DB migration and quick-start example get you typing TypeScript first steps in minutes. Early adopters dig the no-env-reads policy and validation that prevents embedding mismatches.

Who should use this?

TypeScript devs building AI agents—think solo hackers prototyping chatbots with persistent recall, or teams in GitHub repos needing semantic search over user prefs without LangGraph overhead. Ideal for agents in TypeScript GitHub Actions or pages, where you want Postgres-backed memory but hate Python bridges.

Verdict

At 11 stars and 1.0% credibility, it's pre-1.0 alpha—solid docs and tests, but light on features like episodic memory. Grab it for agent prototypes if Postgres fits; contribute to the roadmap if you need more.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.