Unlock OpenClaw’s Power: Step‑by‑Step Guide to Installing and Using Skills

Prerequisites

Run an OpenClaw instance. Deploy it using the one‑click guide at https://cloud.tencent.com/developer/article/2624003.

Default integrated Skills

The latest OpenClaw template includes the agent‑browser v0.2.0 plugin, which enables an AI agent to control a browser (open pages, click elements, input text, capture screenshots).

First query the available Skills (e.g., “现在有哪些可用的Skills”) because the AI may not yet know the built‑in Skills.

Example prompts (Chinese):

用浏览器访问下百度，看一下页面内有什么内容

帮我用浏览器查询下今日5大热点新闻

Installing a new Skill (Email example)

Skills are distributed via the Clawhub Skills marketplace at https://www.clawhub.com/skills. To install the Email skill:

Open the Clawhub Skills page and locate the skill named “email”.

Copy the exact skill name.

In the Lighthouse console, open the OpenClaw instance’s application management page.

Switch to the Skills tab, paste the skill name into the input field, and click Install .

After a short wait the skill appears in the list.

Verify and use the new Skill via QQBot

If OpenClaw is connected to a QQBot, refresh the bot’s skill list with the command 更新一下你的技能列表, then query available skills with 现在有哪些可用的Skills. Once the Email skill is listed, you can send or retrieve emails through chat.

Summary of workflow

Deploy OpenClaw.

Use the built‑in agent‑browser skill for web browsing and searching.

Install additional Skills from Clawhub (e.g., Email) via the Lighthouse UI.

Refresh the QQBot skill list and interact with the new Skill using natural‑language prompts.