Install AutoClaw in One Minute: Quick Setup for a Local AI Assistant

AutoClaw wraps the open‑source OpenClaw client, turning a half‑day installation into three simple steps—download, install, and auto‑configure—while adding seamless Feishu integration, support for GLM‑5 and pony‑alpha‑2 models, built‑in skills, and security recommendations for custom skill creation.

Old Zhang's AI Learning
Old Zhang's AI Learning
Old Zhang's AI Learning
Install AutoClaw in One Minute: Quick Setup for a Local AI Assistant

Installation

AutoClaw wraps OpenClaw, providing identical AI capabilities with a simplified three‑step setup: download the package, run the installer, and let the program automatically configure the environment. After these steps the assistant runs locally and can be invoked from Feishu.

Model support

AutoClaw ships with the flagship open‑source GLM‑5 model (described as the world’s strongest open model; see https://mp.weixin.qq.com/s?__biz=MzA4MjYwMTc5Nw==∣=2649009406&idx=1&sn=84ea06c94dfa9b50e663deb6c5eb778b) and a beta model named pony‑alpha‑2, which the author observed to be more stable, faster, and better suited for skill calls, scheduled tasks, and continuous workflows.

Third‑party model compatibility

Because the original OpenClaw quickly consumes tokens (the author notes a daily cost of 30 CNY is insufficient), AutoClaw accepts arbitrary model APIs. Users supply an API Key and Base URL; the author verified that Ollama’s free cloud models can be connected, and local models should work similarly.

Feishu integration

Integration with Feishu is fully automated. Scanning a QR code authorizes the assistant, eliminating the manual steps required by OpenClaw (creating an app, a bot, and configuring permissions). The client also offers a mobile interface that can remotely control AutoClaw.

Built‑in Feishu skills

feishu-doc – read/write Feishu documents

feishu-drive – manage Feishu cloud drive

feishu-wiki – access Feishu knowledge base

feishu-perm – control Feishu permissions

feishu-chat-history – retrieve group chat records

Skill ecosystem and security

The author notes that the Skills subsystem is a high‑risk area; unverified skills may contain vulnerabilities or malicious code. AutoClaw includes many verified skills produced by Zhipu, and provides a skills‑creator tool for generating personal skills. The author prefers using skills‑creator and avoiding third‑party skills that have not been vetted.

Utility tools

Three utilities are highlighted:

find‑skills – search for skills that match user needs.

skills‑creator – generate custom skills; the author’s own skills were built entirely with this tool.

A strong recommendation to refrain from using unverified third‑party skills.

Intelligent browser automation skill

Zhipu’s official browser‑automation skill can perform a range of web actions: opening pages, searching, browsing social media, interacting with elements, logging in, filling forms, taking screenshots, scraping content, price comparison, news reading, and operating online documents.

Personal demonstrations

In the author’s tests:

A custom svg skill generated SVG images, which were then converted to PNG via an svg2png skill.

An audio‑download skill fetched audio from a video and returned it as an MP3 file.

These examples illustrate the convenience of invoking skills from the Feishu client.

Conclusion

The author finds AutoClaw substantially more convenient than the original OpenClaw and considers it indispensable for daily workflow.

Experience URL: https://autoglm.zhipuai.cn/autoclaw/ (supports macOS and Windows).

Code example

模型方面,除了自家[旗舰 GLM-5 大模型](
# 世界上最强大的开源模型,GLM-5 实测,本地部署,Ollama 免费提供云端模型
),还有一个神秘的 pony-alpha-2,据说是
FeishuAI AssistantLocal DeploymentSkillsOpenClawGLM-5AutoClaw
Old Zhang's AI Learning
Written by

Old Zhang's AI Learning

AI practitioner specializing in large-model evaluation and on-premise deployment, agents, AI programming, Vibe Coding, general AI, and broader tech trends, with daily original technical articles.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.