← 返回 Skills 市场
Readwise Article Saver
作者
zz123-awef
· GitHub ↗
· v1.0.0
· MIT-0
127
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install readwise-article-saver
功能描述
Save article links to Readwise Reader with automatic content-based tagging. Specialized handling for WeChat Official Account (微信公众号) articles. Use this skill...
安全使用建议
Before installing: (1) Understand that the skill will automatically fetch article HTML and save/send it to Readwise and to whatever LLM provider you configure for llm-task (e.g., OpenRouter). If that data may contain sensitive content, do not enable automatic saving. (2) The only script-level secret required is READWISE_TOKEN; OPENROUTER_API_KEY is needed only if you want automatic LLM tagging via the platform — you can disable llm-task or not provide that key if you prefer to tag manually. (3) The code is small and readable (no hidden endpoints or obfuscated code), but SKILL.md explicitly tells the agent to act without confirmation; consider removing automatic invocation or requiring confirmation, and ensure exec and llm-task are only allowed for agents you trust. (4) The declared required binary 'curl' is not used by the bundled scripts—this is harmless but sloppy. (5) If you proceed, test with non-sensitive URLs first, and review openclaw.json settings and the two Python scripts to confirm behavior matches your privacy/security requirements.
功能分析
Type: OpenClaw Skill
Name: readwise-article-saver
Version: 1.0.0
The skill is a legitimate tool designed to save web articles and WeChat posts to Readwise Reader. Analysis of save_article.py and update_tags.py shows they strictly interact with the official Readwise API (readwise.io) using provided environment tokens. The SKILL.md instructions correctly guide the agent through a multi-step workflow of fetching, classifying (via llm-task), and tagging content without any evidence of malicious intent, data exfiltration, or prompt injection.
能力评估
Purpose & Capability
The skill's name/description match what the code does: saving URLs to Readwise and producing tags via an LLM. READWISE_TOKEN is required and used by the included Python scripts. OPENROUTER_API_KEY is listed as a required config (and shown in the example openclaw.json) because the SKILL.md expects the platform 'llm-task' to use an OpenRouter provider — but the Python scripts themselves do not reference OPENROUTER_API_KEY. Required binaries list includes python3 (used) and curl (declared but not used). These are minor mismatches but explainable by the skill’s use of the platform-level llm-task plugin.
Instruction Scope
SKILL.md instructs the agent to 'execute these steps immediately without asking for confirmation' whenever a user sends a URL. Runtime actions fetch full HTML (server-side for WeChat articles), call Readwise API (uploads URL/HTML), and send a text_preview to the llm-task tool for classification — which will forward data to your configured LLM provider. That behavior is within the stated purpose but has privacy implications (automatic transmission of article content to external services) and could surprise users if they expect manual confirmation.
Install Mechanism
This is an instruction-only skill with bundled, locally-included Python scripts. There is no external download/install spec, no archive extraction, and no third-party install URL. Risk from install mechanism is low.
Credentials
The scripts legitimately require READWISE_TOKEN to call Readwise. OPENROUTER_API_KEY is required only to enable the platform llm-task provider in the example config (not read by the Python files). The skill asks for two configuration values in openclaw.example.json; that is reasonable for its workflow but may surprise users who expect only a Readwise token. No other secrets are requested. Declaring curl as a required binary is unnecessary given the included Python code.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or global agent settings. It relies on normal autonomous invocation (llm-task and exec) which is the platform default. The main privilege concern is behavioural (automatic saves) rather than installation or persistent elevated privileges.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install readwise-article-saver - 安装完成后，直接呼叫该 Skill 的名称或使用
/readwise-article-saver触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
- Initial release of Readwise Article Saver skill.
- Save any article URL to Readwise Reader with automated content-based tags.
- Specialized support for WeChat Official Account (mp.weixin.qq.com) articles, including server-side fetching to avoid parsing issues.
- Automatic tagging using a fixed taxonomy via LLM, with region/key thinker/company tags safeguarded by clear rules.
- Batch-saving supported: multiple links processed in one message.
- User receives clear success, partial, or failure notifications with brief details.
元数据
常见问题
Readwise Article Saver 是什么？
Save article links to Readwise Reader with automatic content-based tagging. Specialized handling for WeChat Official Account (微信公众号) articles. Use this skill... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 127 次。
如何安装 Readwise Article Saver？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install readwise-article-saver」即可一键安装，无需额外配置。
Readwise Article Saver 是免费的吗？
是的，Readwise Article Saver 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Readwise Article Saver 支持哪些平台？
Readwise Article Saver 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Readwise Article Saver？
由 zz123-awef（@zz123-awef）开发并维护，当前版本 v1.0.0。
推荐 Skills