Spomory logo

Spomory

graph-based MCP memory with an RL-trained forgetting policy

Share on:
Listing image

A personal AI memory engine: HippoRAG-style graph retrieval, a LightRAG-style incremental knowledge graph, and a GRPO-trained memory-management policy. Local-first for developers, or cloud-hosted for memory that follows you across apps.

Spomory — an explainable memory graph, trained to manage itself Spomory Product ▾ Local Cloud & Apps Memory Passport Platforms ▾ VERIFIED MCP CLIENTS Claude Desktop Cursor Codex CLI Doubao (豆包) Hermes Agent Gemini CLI VS Code (Copilot) Zed Continue.dev LibreChat Dify Kimi CLI Qwen (阿里云百炼) Browse all platforms → Pricing Resources ▾ Research API Reference Blog Changelog GitHub ↗ 中文 Get Started Menu Open-core memory engine One memory, any AI you use. Never start over — A shared memory across every AI you use. Say it once in any AI. Come back tomorrow in a different one. Your context follows y

Related listings

I’ve published my learnings on several topics that interests me in visual formats, and this one is the latest one I did ...

High performance policy routed IPsec mesh SD-WAN

share a Windows PC's keyboard and mouse with a Mac over QUIC

a word game where every word changes the board

A GitHub contribution for my newborn's feedings, turned into SaaS

nine AI agents that argue against your stock thesis