Explore live open-source projects and AI models.

Transform notes into interactive Quest Maps and visual Memory Maps using AI Skills, with FSRS-powered spaced repetition.

Privacy-first AI memory layer - Signal for AI Memory. E2EE, local-first, works with Claude, Cursor, and any MCP-compatible AI.

Engram 是一款主打“激进速通”的高强度英语词汇记忆的桌面应用。 它的设计初衷是剥离一切非必要的视觉干扰与游戏化机制，提供一个极简、专注的终端，让学习流程保持直接、稳定和高频，以应对短期内需要极大量扩充词汇储备的备考需求（如 CET-6 突破、雅思前置准备等）。

Arno's Spanish Engram layout is an optimized key layout for touch typing in Spanish based on ergonomic considerations, with a protocol and software for creating new, optimized key layouts in other languages.

Auditable offline edge intelligence for low-cost edge devices, with benchmark evidence and public board proof on ESP32-C3.

pytorch implementation of DeepSeek Engram

AI Intelligence Platform -- knowledge graph + semantic search + reasoning + multi-agent debate in a single binary

Agent memory and personality - save tokens, keep track of conversation stack, and make engineering fun again, all at the same time!

Local-first context compression for AI coding tools. One binary saves 85-93% of redundant tokens across every LLM call.

给AI 注入可切换的专家记忆——不只是知识检索，而是"谁 + 知道什么 + 怎么思考"的完整人设。人工策展 + 模型自主检索，比 RAG 更精准。兼容所有 MCP 客户端（Claude Code/Desktop、Cursor、Windsurf、Codex 等）。把memory 分享给别人，对方的 AI 立刻成为同一个专家。