{"skill":{"slug":"light-memory","displayName":"OpenClaw Memory System","summary":"A lightweight, prompt-driven three-layer memory system for long-term context retention in OpenClaw agents with secure, asynchronous data extraction and Markd...","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":86,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1776092530231,"updatedAt":1776093408432},"latestVersion":{"version":"1.0.0","createdAt":1776092530231,"changelog":"light-memory v1.0.0\n\n- Introduces a three-layer LLM Agent memory architecture (L1/L2/L3) to address long-term context retention issues.\n- Fully prompt-driven, cross-platform, and does not depend on Python/Bash or vector databases.\n- Features atomic one-click installation with safety checks and automatic data backup.\n- Users can explicitly mark key information (\"决策:\", \"洞察:\"); system extracts and archives asynchronously via Cron tasks.\n- Includes commands for installation, migration, and health checks, with detailed reports and data safety measures.\n- Integrates natively with OpenClaw, supporting long-term, robust, and auditable memory for agent workflows.","license":"MIT-0"},"metadata":null,"owner":{"handle":"jason99666","userId":"s1748ws6j7xc1rhvrzhk1b05e584swy7","displayName":"Jason99666","image":"https://avatars.githubusercontent.com/u/67641581?v=4"},"moderation":null}