← 返回 Skills 市场
258
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install youtube-comment-moderator
功能描述
AI-powered YouTube comment moderation. Fetches comments, classifies them (spam, question, praise, hate, neutral, constructive), drafts replies, and deletes s...
安全使用建议
This skill appears to be what it says: a local Python-based YouTube comment moderator that needs your YouTube API key, Google OAuth client credentials (to reply/delete), and a Gemini API key (for classification). Before installing or running it:
- Treat OAuth data as sensitive. If you are ever asked in chat to paste a redirect/callback URL, paste only the single-use code portion if instructed — do not post full callback URLs into public chat. The skill’s docs are inconsistent about this; follow the more conservative approach (only the code) or run the OAuth step locally.
- Limit API key scope and rotate/ revoke keys if you stop using the skill. Restrict the YouTube API key to the YouTube Data API and enable referrer restrictions if possible.
- Run the skill in a controlled environment (dedicated VM/container) and inspect the oauth.json and DB files it creates (defaults: skills/youtube-comment-moderator/oauth.json and data/youtube-comment-moderator/moderator.db). Set file permissions so tokens are not world-readable.
- Because source/homepage are unknown, you may want to audit the shipped Python files locally before running, or run the tool in a sandbox. The code here is readable and matches the documentation, but unknown origin is always a trust consideration.
- If you only need read-only classification, use 'monitor' mode and skip OAuth to avoid giving write/delete privileges to the skill.
If you want, I can: summarize the exact OAuth pieces you must create in Google Cloud, produce a minimal checklist to harden API keys/permissions, or highlight all lines in the code that write tokens or make outgoing requests.
功能分析
Type: OpenClaw Skill
Name: youtube-comment-moderator
Version: 1.0.1
The YouTube Comment Moderator bundle is a legitimate tool for automated channel management using the Gemini and YouTube Data APIs. The code uses only Python standard libraries, reducing supply chain risks, and follows standard OAuth 2.0 flows for authorization. Notably, the SKILL.md instructions include security-conscious guidance for the AI agent, specifically warning it to protect the user's sensitive OAuth exchange codes and avoid logging full callback URLs.
能力评估
Purpose & Capability
Name/description (YouTube comment moderation) matches the actual code and requirements: it uses the YouTube Data API (API key for reads, OAuth client for write actions) and a Gemini/GCP API key for classification. The files implement fetching, classifying, a SQLite DB, and OAuth flows — all expected for this functionality.
Instruction Scope
Runtime instructions largely stay within scope (guiding users to supply API keys, run setup, authorize OAuth). Two minor inconsistencies: (1) some guidance warns users to paste ONLY the OAuth 'code' (to avoid sharing sensitive data) while other parts instruct headless users to paste the full callback URL; (2) the SKILL.md repeatedly shows shell command examples but also instructs agents to walk users conversationally and 'not tell them to run shell commands'. These are usability/security inconsistencies that could lead a user to paste more than intended into chat (exposing the OAuth redirect contents).
Install Mechanism
No install spec; this is instruction + code only. The skill includes Python scripts shipped with the package (no external arbitrary downloads or obscure installers). That lowers install risk.
Credentials
Requested credentials are proportional and expected: YOUTUBE_API_KEY for reads, YT_MOD_CLIENT_ID/SECRET for OAuth write actions, GEMINI_API_KEY for classification. Optional env overrides point to local file paths (DB, oauth, config). No unrelated secrets are requested.
Persistence & Privilege
The skill persists state locally (SQLite DB under data/..., oauth tokens under skills/.../oauth.json). always:false and normal autonomous invocation. It does not request system-wide config changes or modify other skills. Writing tokens/DB is expected for this use case; ensure file permissions are acceptable.
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 youtube-comment-moderator - 安装完成后，直接呼叫该 Skill 的名称或使用
/youtube-comment-moderator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
- Initial public release with agent-driven, conversational setup guide for YouTube comment moderation.
- Added full scripts for fetching, classifying, moderating, and storing YouTube comments.
- Includes detailed OAuth setup instructions and environment variable documentation.
- Moderation supports monitor, approval, and auto modes with AI-powered comment classification.
- All major pipeline scripts (setup, fetch, classify, moderate, db) included.
v1.0.0
- Initial release of YouTube Comment Moderator.
- Classifies YouTube comments (spam, question, praise, hate, neutral, constructive) using Gemini Flash.
- Supports automated replies, spam deletion, and sentiment analysis via YouTube Data API v3 and OAuth.
- Interactive setup and flexible modes: monitor, approval, and auto-moderation.
- Includes SQLite persistence, customizable config, and support for agent-driven/OAuth flows.
- Designed for channel moderation, auto-replies, spam management, and approval pipelines.
元数据
常见问题
YouTube Comment Moderator 是什么？
AI-powered YouTube comment moderation. Fetches comments, classifies them (spam, question, praise, hate, neutral, constructive), drafts replies, and deletes s... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 258 次。
如何安装 YouTube Comment Moderator？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install youtube-comment-moderator」即可一键安装，无需额外配置。
YouTube Comment Moderator 是免费的吗？
是的，YouTube Comment Moderator 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
YouTube Comment Moderator 支持哪些平台？
YouTube Comment Moderator 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 YouTube Comment Moderator？
由 Dhruv Kar（@dhruvkar）开发并维护，当前版本 v1.0.1。
推荐 Skills