Skip to content

Why memsearch?

One Memory, Every Agent

memsearch provides persistent memory plugins for 4 major AI coding agent platforms: Claude Code, OpenClaw, OpenCode, and Codex CLI.

Memories written in one platform are searchable from any other. A conversation in Claude Code becomes available context in OpenClaw, Codex, and OpenCode — no extra setup, no manual export.

Both for Agent Users and Agent Developers

  • Agent Users: install a plugin, get persistent memory. Zero commands to learn, zero manual saving.
  • Agent Developers: a complete CLI and Python API for building memory and harness engineering into your own agents.

One tool, two audiences.

Markdown is the Source of Truth

Inspired by OpenClaw — your memories are plain .md files: human-readable, editable, version-controllable.

Milvus is a shadow index — a derived, rebuildable cache. Delete the index, run memsearch index, and everything is back. The real data never leaves your markdown files.

Architecture

See the Architecture deep dive and Design Philosophy for the full picture.

Features

  • Hybrid Search — BM25 sparse + dense vector + RRF reranking for the best recall
  • Smart Dedup — SHA-256 content hashing skips unchanged content on re-index
  • Live Sync — file watcher auto-indexes changes in real time
  • Progressive Disclosure — 3-layer recall: search → expand → transcript
  • Pluggable Embeddings — ONNX (local, free), OpenAI, Google, Voyage, Ollama