← 返回 Skills 市场
126
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install parse-deodao-shared-link
功能描述
解析"得到"笔记分享链接（支持提取正文和下载图片）。自动提取得到（dedao.cn）分享链接的正文内容，并支持将文章中的图片保存到本地。当用户发送得到笔记分享链接时使用此技能。
安全使用建议
This skill appears to implement a dedao.cn scraper, but it has several red flags you should address before running: (1) run_parse.js writes output to a hard-coded Windows path (D:/notes/biji/0000) instead of the documented ./images or a configurable outputDir — change this to a relative or configurable directory to avoid overwriting files; (2) SKILL.md and test files reference parseDedao/other exports but parse.js only exports parsePage — verify and fix the API surface so you know which functions will run; (3) the skill depends on Playwright/Chromium but includes no install steps — ensure you install Playwright and its browser binaries in a controlled environment; (4) review the code locally to confirm there are no hidden network endpoints or logging to unexpected services (there are none obvious in the supplied files); (5) run the scripts in a sandbox or VM with limited permissions first to confirm behavior and to prevent accidental writes to system drives. Given the inconsistencies, treat this as potentially useful but untrusted until you fix/configure the output paths and verify dependencies.
功能分析
Type: OpenClaw Skill
Name: parse-deodao-shared-link
Version: 1.0.1
The skill bundle exhibits high-risk behaviors including the use of hardcoded absolute Windows file paths (`D:/notes/biji/0000`) in `scripts/run_parse.js` and `scripts/test_parse.js`, which could lead to unintended file overwrites. Additionally, `scripts/parse.js` launches the browser with the `--no-sandbox` and `--disable-setuid-sandbox` flags, significantly weakening the security posture of the browser automation. While these appear to be unintentional design flaws or artifacts of a local development environment rather than intentional malware, they represent significant vulnerabilities in a shared skill bundle.
能力评估
Purpose & Capability
The name/description and most code align: parse a dedao.cn share link, extract text and images using Playwright. However there are incoherences: SKILL.md and examples mention parseDedao and other site-specific exports, but scripts/parse.js only exports parsePage. SKILL.md says images default to ./images/, but run_parse.js writes to a hard-coded Windows path (D:/notes/biji/0000). These mismatches suggest sloppy packaging and unexpected filesystem behavior.
Instruction Scope
SKILL.md describes extracting content and saving images and provides usage examples. The runtime code does that, but run_parse.js will write markdown and image files to an absolute path (D:/notes/...), which is not documented in SKILL.md. The parse implementation fetches pages and image URLs (network I/O) and writes files to disk — appropriate for a scraper but the undocumented hard-coded path and mismatched exported functions broaden scope unexpectedly.
Install Mechanism
There is no install spec (instruction-only plus shipped scripts). SKILL.md lists dependencies (Playwright and Chromium) but the package does not include installation steps. Playwright requires installing browser binaries and can be large; absence of an install mechanism means the environment must already provide these, which may confuse users and lead them to run the scripts in an environment missing required components.
Credentials
The skill requests no secrets or env vars, which is appropriate. However the code writes files to the local filesystem and uses a hard-coded absolute Windows path (D:/notes/biji/0000). This is disproportionate to the description's implied default of './images' and could overwrite or create files in unexpected locations. There are no network exfiltration endpoints in the code, but file-write behavior requires care.
Persistence & Privilege
The skill is not always-enabled and does not modify other skills or agent configs. It does, however, write persistent files to disk when executed (markdown and images). That is expected for this use case, but the hard-coded output location increases the chance of undesired persistence on the host.
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 parse-deodao-shared-link - 安装完成后，直接呼叫该 Skill 的名称或使用
/parse-deodao-shared-link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
v1.0.1
增加下载关键图片
v1.0.0
Initial release.
- Added skill to parse "得到" (dedao.cn) note sharing links.
- Automatically extracts the main content from shared dedao.cn note URLs, supporting dynamic (JavaScript-rendered) pages.
- Utilizes Playwright with Chromium for headless browser-based content extraction.
- Includes fallback selector logic to improve extraction success.
- Provides detailed usage instructions and troubleshooting notes.
元数据
常见问题
Parse Dedao 是什么？
解析"得到"笔记分享链接（支持提取正文和下载图片）。自动提取得到（dedao.cn）分享链接的正文内容，并支持将文章中的图片保存到本地。当用户发送得到笔记分享链接时使用此技能。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 126 次。
如何安装 Parse Dedao？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install parse-deodao-shared-link」即可一键安装，无需额外配置。
Parse Dedao 是免费的吗？
是的，Parse Dedao 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Parse Dedao 支持哪些平台？
Parse Dedao 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 Parse Dedao？
由 ymsha88（@ymsha88）开发并维护，当前版本 v1.0.1。
推荐 Skills