qmd-memory

Community

Local hybrid search, zero memory API costs

Authorxintuchain
Version1.0.0
Installs0

System Documentation

What problem does it solve?

QMD Memory eliminates recurring API costs and latency caused by remote memory and embedding lookups by running hybrid search locally and serving semantic context to OpenClaw agents.

Core Features & Use Cases

  • Local hybrid search combining BM25 (SQLite FTS5), local vector embeddings, and LLM re-ranking to return high-quality context without external API calls.
  • Auto-configured collections & context that index workspace files, daily logs, intelligence, and project documents to make retrieval precise and relevant.
  • Nightly indexing and multi-agent support with optional MCP server to share a single memory index across multiple agents for collaborative workflows.
  • Use case: a team of agents querying past decisions, research, and daily logs for planning meetings while avoiding embedding API charges and keeping data local.

Quick Start

Run the OpenClaw skill setup to install QMD, configure collections, and generate local embeddings for your workspace.

Dependency Matrix

Required Modules

@tobilu/qmdbc

Components

scripts

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: qmd-memory
Download link: https://github.com/xintuchain/tongtong/archive/main.zip#qmd-memory

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.