OpenClaw Docs Guide

OpenClaw 平台完整使用指南。涵盖 Gateway 配置、频道设置（Telegram/Discord/WhatsApp/微信等）、定时任务、会话管理、安全策略、沙盒配置、模型管理、Agent 管理、设备配对、心跳机制、CLI 命令等。当需要：(1) 配置或排查 OpenClaw (2) 添加频道/定时任务/...

Before installing: 1) Ask the publisher to explain and explicitly declare any filesystem paths the skill will read or write (e.g., /mnt/nas/openclaw-docs, MEMORY.md). 2) Refuse or sandbox any skill that auto-updates from mounted network shares without explicit consent. 3) Require the skill to list exact commands it will run and to limit updates to an opt-in/manual mode (disable automatic heartbeats). 4) If you must use it, run in an isolated environment where the agent cannot access sensitive mounts or credentials, and verify that it cannot modify global OpenClaw config unexpectedly. 5) If the author can show the skill only reads a vetted docs directory and only writes to a declared, user-controlled path, that would reduce risk; otherwise treat it as suspicious.

Type: OpenClaw Skill Name: openclaw-docs-guide Version: 1.1.0 The bundle is a comprehensive knowledge base and documentation guide for the OpenClaw platform, containing summaries of core concepts, configuration examples, and a learning history. While SKILL.md and RELEARN_PLAN.md contain instructions for the agent to automate its learning process and update its internal memory (MEMORY.md), these actions are entirely consistent with the stated purpose of a documentation assistant. No indicators of data exfiltration, unauthorized command execution, or malicious prompt injection were found; the referenced file paths (e.g., ~/.openclaw/openclaw.json) and CLI commands are standard for OpenClaw administration.