360 Secure SkillHub
360 安全技能中心

Vector Memory (Windows) for OpenClaw

面向 OpenClaw 的完整记忆工程技能包，含三个模块：LanceDB 语义存储与检索后端、基于 git-notes 的分支感知长期记忆、记忆清理与维护手册。通过命令行完成记忆写入、检索、更新、遗忘、分支合并与会话同步，默认数据目录可用环境变量覆盖。

sea2049
sea2049
热度 85.40
最近更新 2026-05-17 13:44:47
分类：知识依赖：代码执行 / 读写文件应用场景：笔记整理 / 工作流 / 自动化脚本

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add vector-memory-windows
Metadata
name
vector-memory-windows
description
Full memory stack for OpenClaw on Windows. Includes LanceDB semantic memory, git-notes decision memory, and memory hygiene workflow.

Vector Memory (Windows)

Complete memory engineering bundle for OpenClaw.

Included Modules

  • bundle/lancedb-memory: semantic storage and retrieval backend
  • bundle/git-notes-memory: branch-aware git-notes long-term memory
  • bundle/memory-hygiene: cleanup and maintenance playbook

Why use this bundle

  • Works on Windows out of the box.
  • Keeps compatibility with existing memory APIs.
  • Provides both retrieval and governance, not only storage.

Quick Start

  1. Install LanceDB dependencies:
pip install -r bundle/lancedb-memory/requirements.txt
  1. For semantic memory backend, import from:
  • bundle/lancedb-memory/final_memory.py
  • bundle/lancedb-memory/clawdbot_memory.py
  1. For branch-aware memory:
python bundle/git-notes-memory/memory.py -p . sync --start
  1. Apply hygiene recommendations from:
  • bundle/memory-hygiene/SKILL.md

Storage Defaults

  • Windows: D:\clawtest\memory\lancedb
  • Unix-like: ~/.clawdbot/memory/lancedb

You can override with:

  • OPENCLAW_LANCEDB_PATH
  • CLAWTEST_ROOT (Windows)