Webpuppet Rs MCP 伺服器 | PlugMCP
跳到正文 webpuppet-rs-mcp
MCP (Model Context Protocol) server for webpuppet browser automation.
This crate provides a standards-compliant MCP server that exposes webpuppet functionality as tools for AI assistants like GitHub Copilot, Claude Desktop, and other MCP-compatible clients.
MCP server for webpuppet browser automation
把任意工具接入任意代理。面向非開發者的 MCP 伺服器目錄;閘道承載流量,n8n 承載應用。
產品
目錄
開發者
公司
Features
MCP-compliant : Implements JSON-RPC 2.0 over stdio (standard MCP transport)
Tool exposure : Exposes AI prompting, screenshot, and research capabilities
Security guardrails : Inherits webpuppet's permission system
Response screening : Filters prompt injections and malicious content
Browser detection : Automatically finds Brave/Chrome/Chromium
Human intervention : Pause/resume for captchas, 2FA, and manual steps
Available Tools
Installation # Build and install
cargo install --path .
# Or run from source
cargo run -p webpuppet-mcp -- --stdio
Usage with VS Code / GitHub Copilot Add to your .vscode/mcp.json:
{
"servers": {
"webpuppet": {
"command": "webpuppet-mcp",
"args": ["--stdio"],
"env": {}
}
}
}
Or if running from cargo:
{
"servers": {
"webpuppet": {
"command": "cargo",
"args": ["run", "-p", "webpuppet-mcp", "--", "--stdio"],
"cwd": "/path/to/webpuppet",
"env": {}
}
}
}
Usage with Claude Desktop Add to your claude_desktop_config.json:
{
"mcpServers": {
"webpuppet": {
"command": "webpuppet-mcp",
"args": ["--stdio"]
}
}
}
Human Intervention System The MCP server includes tools for human-in-the-loop workflows:
When Intervention is Needed
Captcha : reCAPTCHA, hCaptcha, Cloudflare challenges
Two-Factor Auth : SMS codes, TOTP, email verification
Login : Session expired, auth required
Rate Limits : Too many requests
Workflow
Agent calls webpuppet_intervention_status to check state
If intervention needed, agent notifies user
User completes manual task in visible browser
User/agent calls webpuppet_intervention_complete with success=true
Automation resumes
Example Agent: "I need to send a prompt to Claude but see a captcha..."
Agent: [calls webpuppet_intervention_status]
Agent: "⚠️ A captcha is displayed. Please complete it in the browser."
User: [solves captcha manually]
User: "Done!"
Agent: [calls webpuppet_intervention_complete with success=true]
Agent: "Thank you! Continuing..."
Dependency: security-mcp wrap (DRAFT) Recommended fleet layout for IDE/cabal agents:
Example (operator-owned; validate paths and tokens locally):
{
"servers": {
"webpuppet": {
"command": "security-mcp",
"args": ["--stdio", "--wrap", "--wrap-command", "webpuppet-mcp", "--wrap-arg", "--stdio"],
"env": {
"SECURITY_MCP_WRAP": "1"
}
}
}
}
Do not treat wrap as a pinned STABLE contract until security-mcp bulletin promotion
and consumer acknowledgment in this repo are recorded (separate PR).
Security Model All operations are subject to the webpuppet permission system:
Default (Secure) Policy
✅ Allowed : Navigate, ReadContent, SendPrompt, ReadResponse, NewConversation, ContinueConversation, Screenshot
❌ Blocked : DeleteAccount, ChangePassword, ModifyPayment, RevokeTokens, FileSystemAccess, etc.
🌐 Domains : Only AI provider domains (claude.ai, x.com, gemini.google.com)
🌐 HTTPS-only : In secure mode, http:// URLs are denied
⚠️ Risk Threshold : Max risk level 5 (out of 10)
Permission Policies # Secure (default) - blocks destructive operations, allows AI interaction
webpuppet-mcp --policy secure
# Read-only - only allows reading, no prompts or modifications
webpuppet-mcp --policy readonly
# Permissive - allows most non-destructive operations (use with caution)
webpuppet-mcp --policy permissive
Response Screening All AI responses are automatically screened for:
Invisible text : Zero-width characters, 1pt fonts
Prompt injections : "Ignore previous instructions" patterns
Encoded payloads : Base64, hex encoded content
Hidden elements : CSS display:none, opacity:0
If screening detects issues, the response is sanitized and a warning is included.
Example Tool Calls
Send a Prompt {
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "webpuppet_prompt",
"arguments": {
"provider": "claude",
"message": "Explain how io_uring works in Rust",
"context": "Focus on memory safety"
}
}
}
Check Permission {
"jsonrpc": "2.0",
"id": 2,
"method": "tools/call",
"params": {
"name": "webpuppet_check_permission",
"arguments": {
"operation": "DeleteAccount"
}
}
}
{
"jsonrpc": "2.0",
"id": 2,
"result": {
"content": [{
"type": "text",
"text": "# Permission Check\n\n**Operation**: `DeleteAccount`\n**Status**: ❌ DENIED\n**Reason**: Operation explicitly denied by policy\n**Risk Level**: 10/10"
}],
"isError": false
}
}
Architecture ┌─────────────────────────────────────────────────┐
│ MCP Client │
│ (VS Code, Claude Desktop, etc.) │
└─────────────────────┬───────────────────────────┘
│ JSON-RPC 2.0 / stdio
┌─────────────────────▼───────────────────────────┐
│ webpuppet-mcp Server │
│ ┌───────────────────────────────────────────┐ │
│ │ Permission Guard │ │
│ │ - Operation allowlist/blocklist │ │
│ │ - Domain restrictions │ │
│ │ - Risk level enforcement │ │
│ │ - Audit logging │ │
│ └───────────────────────────────────────────┘ │
│ ┌───────────────────────────────────────────┐ │
│ │ Tool Registry │ │
│ │ - webpuppet_prompt │ │
│ │ - webpuppet_screenshot │ │
│ │ - webpuppet_list_providers │ │
│ │ - webpuppet_detect_browsers │ │
│ │ - webpuppet_check_permission │ │
│ └───────────────────────────────────────────┘ │
└─────────────────────┬───────────────────────────┘
│
┌─────────────────────▼───────────────────────────┐
│ webpuppet │
│ - Browser automation (Brave/Chrome) │
│ - AI provider integration │
│ - Content screening │
│ - Session management │
└─────────────────────────────────────────────────┘
License
Status & roadmap Scrapling 🕷️ The BEST Web Scraping framework that handles everything from a single request to a full-scale crawl! Don't be shy, join here: https://discord.gg/EMgGbDceNQ and follow here for daily tips and trick Chrome Devtools Chrome DevTools for coding agents Chrome DevTools MCP MCP server for Chrome DevTools Chrome DevTools MCP MCP server for Chrome DevTools Chrome Devtools Chrome DevTools for coding agents UI TARS Desktop The Open-Source Multimodal AI Agent Stack: Connecting Cutting-Edge AI Models and Agent Infra Playwright Playwright MCP server Playwright Playwright MCP server Obscura The headless browser for AI agents and web scraping Cua Scale computer-use 2.0 with open-source drivers, cross-OS sandboxes, and benchmarks for training, evaluation, and data generation. Ego Lite The fastest browser for AI agents to run browser automation, built for sharing your logged-in browser state with your AI agents, like Codex or Claude Code, without disturbing you. Zero cost, zero conf SeleniumBase 📊 Python APIs for Web Automation, Testing, and Scraping. CDP Mode launches a stealth browser that bypasses bot-detection/CAPTCHAs. Adds stealth to Playwright and has an MCP server for AI agents. Browser automation MCP 伺服器 最佳 Browser automation MCP 伺服器
尚無工具清單。遠端伺服器會在首次健康檢查後取得工具清單。
如何安裝 Webpuppet Rs? 在本頁複製適用於 Cursor、Claude Desktop、Claude Code 或 VS Code 的 stdio 設定。它使用發布者的安裝方式,而不是 PlugMCP 代理伺服器。
Webpuppet Rs 需要 API 金鑰嗎? 這筆記錄未列出必要的環境變數。
Webpuppet Rs 使用哪種傳輸方式? Webpuppet Rs 透過 stdio 在本機執行。
Webpuppet Rs 上次檢查是什麼時候? 上次檢查時間:2026-10-28。 MCP server for webpuppet browser automation
工具
0 / 10