据 AlphaMoat 统计,截至 2026-09,Bits Browser Automation 累计下载 3,377 次、获 0 个星标,在全部 5,078 个 AI 技能中排名第 3,331 位,在「AI 智能体」分类(969 个)中排名第 681 位。
| 平台 | 累计安装 | 累计下载 | 星标 | 评论 |
|---|---|---|---|---|
| ClawHub ↗ | 106 | 3,377 | - | - |
| SkillHub ↗ | 105 | 3,329 | - | — |
各平台的计数口径不同,统计时间不一致。
主文介绍网页导航、页面读取、结构化提取、表单交互、OAuth 流程和多窗口操作,并通过沙箱中的 TypeScript SDK 运行任务。配置要求 API key,安装示例使用 npx 下载 MCP 包;浏览器可能处于已登录状态,任务也可能把数据发送到 Bits、目标网站或执行包。使用前应固定并审阅软件版本,采用专用且可撤销的密钥;未经审查不要提交敏感或受监管数据,并在执行写操作前明确确认目标和内容。
资料发布于 2026-10-29
以下是根据技能文档生成的示例,不代表实际运行结果。
2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。
| 月份 | 累计下载 | 环比 | 星标 |
|---|---|---|---|
| 2026-09 | 3,377 | — | - |
| 2026-08 | 3,246 | +5.2% | - |
| 2026-07 | 3,086 | +10.9% | - |
| 2026-06 | 2,782 | — | - |
| 2026-03 | 2,045 | — | - |
以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。
--- name: bits-mcp description: Control browser automation agents via the Bits MCP server. Use when running web scraping, form filling, data extraction, or any browser-based automation task. Bits agents can navigate websites, click elements, fill forms, handle OAuth flows, and extract structured data. --- # Bits MCP - Browser Automation Bits is an AI browser automation platform. The MCP server lets you run browser automation tasks from your AI assistant. ## Setup
### 1. Get an API Key
1. Go to [app.usebits.com](https://app.usebits.com)
2. Sign in with Google
3. Navigate to **Settings → API Keys**
4. Click **Create API Key**, give it a name
5. Copy the key (starts with `bb_`) — you won't see it again
### 2. Configure MCP
Add to your MCP config (e.g., `~/.openclaw/openclaw.json`):
```json
{
"mcpServers": {
"bits": {
"command": "npx",
"args": ["-y", "usebits-mcp"],
"env": {
"BITS_API_KEY": "bb_your_key_here"
}
}
}
}
```
For Claude Code (`~/.claude.json`):
```json
{
"mcpServers": {
"bits": {
"command": "npx",
"args": ["-y", "usebits-mcp"],
"env": {
"BITS_API_KEY": "bb_your_key_here"
}
}
}
}
```
### 3. Restart
Restart your gateway/client to pick up the new MCP server.
## Usage
The Bits MCP uses "Code Mode" — you write TypeScript SDK code that executes in a sandbox. Two tools are available:
1. **Documentation search** — Query the SDK docs
2. **Code execution** — Write and run TypeScript against the Bits SDK
### Example: Scrape a Website
```
Use the Bits MCP to go to news.ycombinator.com and get the top 5 story titles
```
The agent will:
1. Search docs for navigation/scraping methods
2. Write TypeScript code to navigate and extract data
3. Execute it and return results
### Example: Fill a Form
```
Use Bits to go to example.com/contact, fill out the contact form with name "Test" and email "[email protected]", then submit
```
### Example: Extract Structured Data
```
Use Bits to scrape the product listings from example-store.com/products and return them as JSON with name, price, and URL fields
```
## Capabilities
- **Navigate** — Go to URLs, handle redirects
- **Read pages** — Extract text, get page layouts, take screenshots
- **Interact** — Click elements, fill inputs, press keys
- **Handle auth** — OAuth popups, login forms, 2FA (with stored credentials)
- **Multi-window** — Switch between tabs/popups
- **Structured output** — Return data in specific JSON schemas
## Creating Workflows (Optional)
For repeated tasks, create a workflow in the Bits web app:
1. Go to [app.usebits.com](https://app.usebits.com) → **Workflows**
2. Create a workflow with a definition (instructions for the agent)
3. Optionally add an output schema for structured responses
4. Run via API: `POST /workflows/{id}/runs`
## Troubleshooting
**"API key invalid"** — Check your key starts with `bb_` and is copied correctly.
**Slow startup** — First run downloads the MCP package via npx. Subsequent runs are faster.
**Task stuck** — Browser automation can hit CAPTCHAs or unexpected modals. Check the live view URL in the response.
## Links
- Web app: [app.usebits.com](https://app.usebits.com)
- API docs: [api.usebits.com/openapi.json](https://api.usebits.com/openapi.json)
把开发过程中的经验、错误与纠正记录为 Markdown 文件,供后续复用并晋升为工作区记忆。
用自然语言描述需求,在本地与远程技能市场里找技能并给出推荐的技能。
给 AI agent 用的浏览器自动化命令行技能。
把用户纠正与自我复盘沉淀成分层记忆、并按使用情况升降级的技能。
在安装第三方技能之前,按清单审查来源、权限与危险模式的技能。
把 agent 的记忆与跨技能状态存成带类型和约束的知识图谱的技能。
一份为 AI 代理规划主动跟进、工作记忆与复盘流程的工作区配置指南。
指导如何创建与更新有效 Claude 技能的写作指南。
数据月份:2026-09 · 下载/星标/安装数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。