Dingtalk

Extract and summarize titles, authors, updates, and content from publicly shared Dingtalk pages, apps, and group invitations without internal API access.

Security Scan

The skill's claims and runtime instructions are internally consistent: it only targets publicly shared Dingtalk pages, requests no credentials, has no install steps, and stays within the stated scope.

Purpose & Capability

Name/description (extract and summarize public Dingtalk pages) match the SKILL.md: it describes extracting titles, authors, timestamps, content summaries and explicitly limits itself to publicly accessible pages without internal API access.

Instruction Scope

Instructions stay within scope (fetch and parse public share/app/invite pages). The doc notes 'content is dynamically loaded, wait for rendering' which implies a need for a JS-capable fetch/render step (headless browser or renderer). The skill does not request browser binaries or list how rendering should occur — this is a mild operational gap but not a security incoherence.

Install Mechanism

No install spec or code files are present (instruction-only), so nothing is written to disk and no external packages are pulled.

Credentials

No environment variables, credentials, or config paths are required. This is appropriate given the stated purpose of only handling public pages.

Persistence & Privilege

always is false and model invocation is normal; the skill does not request persistent/privileged presence or modify other skills' settings.

Assessment

This skill appears coherent and limited to public Dingtalk content, but note: it will require network access and a way to render JavaScript-driven pages (e.g., a headless browser) which the platform/agent must provide. Before installing, confirm the agent environment can safely perform web rendering and scraping, obey robots.txt and Dingtalk's terms of service, and implement rate limiting to avoid being blocked. The skill does not request credentials — do not provide any enterprise tokens; if later you are asked for auth, treat that as a red flag.

钉钉

用途与边界

• 面向公开应用与分享页面的信息提取与摘要
• 不提供企业内部API调用或权限绕过能力
• 仅处理公开可访问页面

关键入口

• 主页：https://www.dingtalk.com/
• 公开应用页/公告页：由分享链接进入

常见任务

• 公告分享页的标题/作者/更新时间与正文摘要
• 公开应用页的功能说明、版本信息与链接整理
• 公开群邀请页的基本说明与注意事项摘要

数据字段

• 标题、作者、更新时间、正文摘要、标签、来源链接
• 应用名称、功能概述、版本信息、链接
• 群邀请页描述与条款摘要

自动化要点

• 内容动态加载，需等待渲染完成后解析
• 无企业内部鉴权能力，仅处理公开分享
• 频率控制避免重复访问

示例流程

• 公告摘要：访问分享页 → 解析标题/作者/更新时间 → 抽取正文摘要
• 应用整理：进入应用页 → 抽取功能/版本信息 → 输出链接清单
• 群页面摘要：访问邀请页 → 抽取说明与条款 → 输出注意事项

合规提示

• 遵守平台与企业信息安全规定，不处理非公开内容
• 输出仅用于内部分析与提醒