AI品牌管理工具

BNBot是一个AI驱动的个人品牌管理工具,用于跨平台内容创建、趋势发现和增长自动化。

  • by bnbot-ai
  • TypeScript
  • 本地部署
  • 更新于
在 GitHub 查看 bnbot-ai/bnbot-mcp-server

使用说明

@bnbot/cli

BNBot(品牌&机器人) --X的人工智能个人品牌工具包。从30多个平台发现趋势,用人工智能创建内容,并自动化您的增长。

快速开始

npx @bnbot/cli setup

这将全局安装CLI并设置Claude Code技能(/bnbot).

然后安装 BNBot Chrome扩展程序 用于浏览器自动化。

建筑

Terminal
    -> bnbot x post "Hello!"
    | WebSocket (ws://localhost:18900)
BNBot Chrome Extension
    | chrome.scripting / DOM / Internal APIs
X + 30 platforms (TikTok, YouTube, Reddit, Bilibili, etc.)

命令

设置和状态

bnbot setup                     # Install CLI + Claude skill
bnbot serve                     # Start WebSocket server
bnbot login                     # Login to BNBot
bnbot status                    # Check extension connection

X平台——推特和互动

bnbot x post "Hello!"                          # Post a tweet
bnbot x post "Look" -m ./photo.png             # Post with media
bnbot x post "Draft" -d                        # Draft mode
bnbot x thread '[{"text":"1"},{"text":"2"}]'   # Post a thread
bnbot x reply <url> "Great post!"              # Reply to a tweet
bnbot x quote <url> "My thoughts"              # Quote tweet
bnbot x like <url>                             # Like a tweet
bnbot x unlike <url>                           # Unlike
bnbot x retweet <url>                          # Retweet
bnbot x unretweet <url>                        # Unretweet
bnbot x follow elonmusk                        # Follow a user
bnbot x unfollow elonmusk                      # Unfollow
bnbot x delete <url>                           # Delete a tweet
bnbot x bookmark <url>                         # Bookmark
bnbot x unbookmark <url>                       # Unbookmark
bnbot x close                                  # Close composer

X平台--报废

bnbot x scrape timeline                        # Scrape home timeline
bnbot x scrape timeline -l 50                  # With limit
bnbot x scrape bookmarks                       # Scrape bookmarks
bnbot x scrape search "AI agents"              # Search tweets
bnbot x scrape search "AI" -t latest -l 50     # Latest tab, 50 results
bnbot x scrape search "AI" --from user --since 2025-01-01
bnbot x scrape user-tweets elonmusk            # User's tweets
bnbot x scrape user-profile elonmusk           # User profile
bnbot x scrape thread <url>                    # Scrape a thread
bnbot x analytics                              # Account analytics

X平台--导航

bnbot x navigate <url>                         # Navigate to URL
bnbot x navigate search "AI agents"            # Go to search
bnbot x navigate bookmarks                     # Go to bookmarks
bnbot x navigate notifications                 # Go to notifications

公共数据抓取器(无需扩展)

bnbot hackernews search "AI agents" -l 20      # Hacker News
bnbot stackoverflow search "react hooks"       # Stack Overflow
bnbot wikipedia search "ChatGPT" --lang zh     # Wikipedia
bnbot apple-podcasts search "tech"             # Apple Podcasts
bnbot substack search "AI newsletter"          # Substack
bnbot v2ex hot                                 # V2EX hot topics
bnbot bloomberg news                           # Bloomberg
bnbot bbc news                                 # BBC
bnbot sinafinance news --type 1                # Sina Finance 7x24
bnbot sinablog search "AI"                     # Sina Blog
bnbot xiaoyuzhou podcast <id>                  # Xiaoyuzhou FM

内容获取(通过扩展)

bnbot weixin article <url>                     # WeChat article
bnbot tiktok fetch <url>                       # TikTok video
bnbot xiaohongshu fetch <url>                  # Xiaohongshu note

向后兼容

传统的kebab case命令仍然有效:

bnbot post-tweet --text "Hello"                # -> bnbot x post "Hello"
bnbot scrape-timeline --limit 10               # -> bnbot x scrape timeline -l 10
bnbot search-hackernews --query "AI"           # -> bnbot hackernews search "AI"

登录

# Auto-login via clawmoney API key
bnbot login

# Or via email
bnbot login --email you@example.com

完整命令参考

看 docs/COMMANDS.md 获取包含所有选项的完整命令参考。

链接

许可证

麻省理工学院