aitytech

Persistent memory system for AI coding assistants via MCP. Works with Claude Code, Cursor, Copilot, Windsurf, Cline.

53
4
100% credibility
Found Feb 03, 2026 at 18 stars 3x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Local persistent memory system for AI coding assistants that stores project decisions, patterns, errors, and context in SQLite for use across sessions via MCP tools.

How It Works

1
🔍 Discover a smarter coding helper

You hear about a tool that helps your AI coding buddy remember decisions and patterns from past work sessions.

2
📦 Install with one simple command

Run a quick setup command in your project folder and it automatically configures everything for your AI tool.

3
🧠 Your AI now has memory

Next time you code, your AI recalls past choices, errors fixed, and project patterns without you repeating yourself.

4
👀 Peek at saved memories anytime

Open a web page or terminal view to browse, search, and manage what your AI has remembered.

5
💾 Save important notes easily

Quickly jot down key decisions or tips using a simple command, tagged for easy finding later.

AI coding feels magical

Sessions build on each other—your AI gets smarter over time, saving you hours of re-explaining context.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 53 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 agentkits-memory?

Agentkits-memory is a TypeScript package delivering persistent memory for AI coding assistants, storing decisions, patterns, errors, and session context in a local SQLite database. It integrates via the MCP protocol, letting tools like Claude Code, Cursor, Windsurf, Cline, and Copilot recall history across sessions without cloud dependency or API keys. Run `npx @aitytech/agentkits-memory` for zero-config setup, unlocking CLI saves, vector search, and a web viewer at localhost:1905.

Why is it gaining traction?

Unlike scattered markdown notes or cloud-reliant alternatives, it offers github persistent storage that's 100% local, token-efficient with 3-layer search (saving 70% tokens), and multi-platform from one command. Auto-hooks capture tool use and summaries in supported editors, while background workers handle AI enrichment and compression. Solid 91% test coverage and multilingual embeddings (100+ languages) make it reliable for persistent memory llm workflows.

Who should use this?

Backend devs building apps with Cursor or Claude Code, where AI repeats auth patterns or error fixes. Frontend teams using Windsurf or Cline on shared projects needing consistent context. Any agentkits user tired of re-explaining project conventions across sessions.

Verdict

Promising for MCP-based AI assistants, with excellent docs and 970 passing tests, but 32 stars and 1.0% credibility score signal early maturity—expect some rough edges. Try it if you're in the ecosystem; skip for production without more adoption. (198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.