kevin-hs-sohn/hipocampus
Drop-in memory harness for AI agents — 3-tier memory, compaction tree, hybrid search. One command to set up. Works with Claude Code and OpenClaw.
Maintains a ~3K token topic index (ROOT.md) auto-loaded each session that compresses conversation history into a scannable overview, letting agents recognize cross-domain connections without explicit search. Uses a self-compacting hierarchical tree (raw → daily → weekly → monthly → root) that applies LLM summaries only above configurable thresholds, preserving detail while managing token growth. Integrates directly into Claude Code and OpenCode via plugin architecture, with hybrid search combining BM25 and optional vector search (qmd) for both keyword and semantic retrieval.
12 stars and 3,942 monthly downloads. Available on npm.
Stars
12
Forks
1
Language
JavaScript
License
MIT
Category
Last pushed
Mar 17, 2026
Monthly downloads
3,942
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/kevin-hs-sohn/hipocampus"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Related tools
MemoriLabs/Memori
SQL Native Memory Layer for LLMs, AI Agents & Multi-Agent Systems
volcengine/OpenViking
OpenViking is an open-source context database designed specifically for AI Agents(such as...
zjunlp/LightMem
[ICLR 2026] LightMem: Lightweight and Efficient Memory-Augmented Generation
mem0ai/mem0
Universal memory layer for AI Agents
memodb-io/memobase
User Profile-Based Long-Term Memory for AI Chatbot Applications.