BigHugger
GH Repository · zilliztech

memsearch

A persistent, unified memory layer for all your AI agents (e.g. Claude Code, Codex, DSH), backed by Markdown and Milvus.

stars
2,614
30-day movement
+207/day
Related entries
60
Connections
2
python/uvdshdeepseek-harnessdeepseekcodexskillspythonopencodePythonagent-memoryclaude-codedsh-pluginsemantic-searchclaude-code-pluginmemoryagentopenclawmilvusragembeddingslong-term-memoryai-agentshybrid-search

memsearch is a persistent, unified memory layer for AI agents such as Claude Code, Codex, and DSH. It stores memory as Markdown files and uses Milvus for embedding-based retrieval, supporting semantic and hybrid search.

It gives your coding agents a shared long-term memory instead of session-scoped context.

Use it to

  • Persist long-term memory across agent sessions
  • Index Markdown notes with Milvus embeddings
  • Run semantic and hybrid search over stored memories
  • Share one memory layer across multiple agents
  • Install as a Claude Code plugin

For Developers running AI coding agents who need persistent memory

Role
rag
Language
Python
Licence
MIT
Forks
251
Open issues
34
Last push
2026-09-12
Latest release
v0.1.16 · 2026-03-09
topicsagent-memoryragmilvussemantic-searchclaude-codeembeddings