Zero-dependency, local-first Agent memory system: 4-arm recall + nebula visualization panel + MCP
Translated from Chinese by Claude ยท original on GitHub โ
Zero-dependency, local-first Agent memory system: 4-arm recall + nebula visualization panel + MCP
Translated from Chinese by Claude ยท original on GitHub โ
// momentum
// why it's here
// how to help
// readme
Zero dependencies ยท Pure local ยท Visual Agent memory system. Ready to use right after installation, free, no cloud.
A self-contained memory engine: local JSONL four-arm recall (lexical / activation / graph diffusion / profile) + an interactive memory nebula visualization panel, with an optional Hindsight HTTP API layer for semantic enhancement. The core engine always runs standalone: no network, no cost, no dependency on any database or LLM.
๐ Beginner who wants to start right away? See the Quick Start (step-by-step tutorial): no coding required, just follow along and youโll be using it.
๐ฅ Using DeepSeek Harness (DSH)? Send this repository URL to DSH and have it install for you (see Install to DSH).
Real-time animation of the memory nebula panel: nebula particles breathe ยท recall cards light up in rotation ยท four-arm hits ยท data chain refreshes live
Works 100% offline. Zero dependencies. No API keys required. 100% offline ยท zero dependencies ยท no API key needed.
Health score
Is the project alive and maintained right now? A 0โ100 pulse recomputed nightly from commit recency, rhythm, how fast issues close, and how quickly PRs get merged.
How it's calculatedUndervalued score
Activity relative to attention (stars + watchers). Above 50 means a lot of activity for a project its size. Recomputed nightly from commit velocity, contributor effort, issue resolution, fork utility, release cadence, and project maturity โ divided by a logarithmic reach factor.
How it's calculated