memsearch and openclaw-engram

Memsearch is a generalized memory library that inspired OpenClaw's specialized memory plugin, making them ecosystem siblings where the plugin implements domain-specific memory capabilities built on concepts pioneered by the broader framework.

memsearch
64
Established
openclaw-engram
52
Established
Maintenance 25/25
Adoption 10/25
Maturity 11/25
Community 18/25
Maintenance 13/25
Adoption 5/25
Maturity 18/25
Community 16/25
Stars: 846
Forks: 77
Downloads:
Commits (30d): 110
Language: Python
License: MIT
Stars: 13
Forks: 6
Downloads:
Commits (30d): 0
Language: TypeScript
License: MIT
No Package No Dependents
No risk flags

About memsearch

zilliztech/memsearch

A Markdown-first memory system, a standalone library for any AI agent. Inspired by OpenClaw.

Implements semantic search over markdown files using pluggable embedding providers (ONNX, Google, Voyage, Ollama, local), with automatic file watching and SHA-256 dedup to skip re-embedding unchanged content. Stores vectors in a local database and exposes a simple async Python API that integrates seamlessly with LLM frameworks like OpenAI, Anthropic Claude, and Ollama for agent-driven recall-think-remember loops.

About openclaw-engram

joshuaswarren/openclaw-engram

Local-first memory plugin for OpenClaw AI agents. LLM-powered extraction, plain markdown storage, hybrid search via QMD. Gives agents persistent long-term memory across conversations.

Engram integrates as a native OpenClaw plugin and MCP server, supporting both cloud (OpenAI) and local LLM-powered extraction (Ollama, LM Studio) with zero API dependencies. Memories persist as git-friendly markdown files with YAML frontmatter and lifecycle management (fact, decision, preference, correction, entity tracking), using hybrid search (BM25 + vector reranking via QMD) to surface contextual knowledge at conversation start. The architecture uses a three-phase recall-buffer-extract pipeline triggered by conversation turns, enabling semantic memory injection across multiple agent harnesses and MCP clients (Claude Code, Codex CLI) on a single machine or distributed setups.

Related comparisons

Scores updated daily from GitHub, PyPI, and npm data. How scores work