← Back to Skills Marketplace
fanyur-wang

Workswith Claw

by Fanyur · GitHub ↗ · v1.0.3 · MIT-0
cross-platform ⚠ suspicious
225
Downloads
0
Stars
0
Active Installs
4
Versions
Install in OpenClaw
/install workswith-claw
Description
为家,注入灵魂。独立于 Home Assistant 的智能家居中间件,通过 HA API 实现设备的语义化理解、习惯学习和智能预判。
Capability Analysis
Type: OpenClaw Skill Name: workswith-claw Version: 1.0.3 The skill bundle acts as a smart home middleware that requires a Home Assistant (HA) long-lived access token and possesses the high-risk capability of writing files to the host filesystem. Specifically, `src/api/routes/apply.py` writes YAML automation files directly into the sensitive `~/.homeassistant/automations` directory. While this behavior is documented in `SKILL.md` as a core feature for habit-based automation, the ability to modify HA configurations and the handling of sensitive credentials (HA tokens and LLM API keys in `src/services/llm_enhancer.py`) represent a significant attack surface. There is no evidence of explicit malice, but the broad filesystem access and external API communication justify a suspicious classification.
Capability Assessment
Purpose & Capability
The skill's stated purpose is a local HA middleware and the code implements HA API clients, device analysis, habit learning, and YAML automation generation — that is coherent. However the registry/metadata claims no required env vars or primary credential while the code and SKILL.md clearly require HA_URL and HA_TOKEN (and optionally WORKSWITH_CLAW_API_KEY). The metadata omission is an incoherence: a HA-integration skill should declare HA credentials in required envs.
Instruction Scope
SKILL.md instructs the user to provide HA_URL and HA_TOKEN and to run the service locally; the code reads HA_TOKEN/HA_URL from .env and will call the Home Assistant REST API and write YAML files into ~/.homeassistant/automations. Those behaviors match the purpose, but there are two problems: 1) authentication is optional — if WORKSWITH_CLAW_API_KEY is not set the API skips auth (dev mode), and the app enables CORS allow_origins=["*"] — meaning the service can be reachable and used without auth if misconfigured; 2) some service names (openclaw_client, llm_enhancer) suggest potential external network activity contrary to SKILL.md's
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install workswith-claw
  3. After installation, invoke the skill by name or use /workswith-claw
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.3
workswith-claw v1.0.3 - 更新了 README.md 文档内容 - 无功能、依赖或接口变更,仅文档调整
v1.0.2
修复:声明HA凭证需求,说明写入automation行为
v1.0.1
修复权限配置,明确与HA关系
v1.0.0
初始版本发布
Metadata
Slug workswith-claw
Version 1.0.3
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 4
Frequently Asked Questions

What is Workswith Claw?

为家,注入灵魂。独立于 Home Assistant 的智能家居中间件,通过 HA API 实现设备的语义化理解、习惯学习和智能预判。 It is an AI Agent Skill for Claude Code / OpenClaw, with 225 downloads so far.

How do I install Workswith Claw?

Run "/install workswith-claw" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Workswith Claw free?

Yes, Workswith Claw is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Workswith Claw support?

Workswith Claw is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Workswith Claw?

It is built and maintained by Fanyur (@fanyur-wang); the current version is v1.0.3.

💬 Comments