bugezy
Voice-powered bug reporting MCP server with 13 tools. Captures console logs, network errors, DOM traces, and developer voice descriptions. AI reads structured reports to find and fix bugs automatically. Chrome Extension + Python CLI.
README
🐛 BugEzy — Voice-Powered Bug Reporting for Developers
Capture bugs by talking. Let AI fix them.
BugEzy is a Chrome extension + MCP server that lets developers report bugs using voice. It captures console logs, network errors, DOM traces, and your voice description automatically — so AI can find the root cause and give you the fix.
🌐 bugezy.dev · 🧩 Chrome Web Store · 📖 SKILL.md (AI Guide)
How It Works
- 🎙️ Hit Record — Open BugEzy popup, hit record. Talk about the bug while operating.
- 📋 Auto-organized — BugEzy captures screen replay, console logs, network errors, and action timeline into a structured report.
- 🤖 AI Fixes It — Hand the report to any MCP-compatible AI assistant. AI finds root causes and gives you the fix.
MCP Server (13 Tools)
Connect your AI assistant to BugEzy:
MCP Endpoint: https://bugezy.dev/mcp (Streamable HTTP)
| Tool | Description |
|---|---|
get_report_overview |
Report metadata + AI bug navigation summary |
get_timeline |
Complete event timeline (console + network + voice + markers) |
get_console_logs |
Console logs (warn/error) |
get_network_errors |
Network errors (4xx/5xx) |
get_voice_transcript |
Developer voice transcript |
get_screenshots |
Report screenshots |
get_rrweb_summary |
DOM trace summary |
get_rrweb_events |
Full DOM events |
get_page_info |
Page info (URL/title/browser/resolution) |
get_metadata |
Custom metadata via SDK |
list_reports |
List user's bug reports |
get_live_errors |
Live console/network errors |
get_terminal_logs |
Terminal error logs |
MCP Config (Claude Desktop / Cursor / Windsurf)
{
"mcpServers": {
"bugezy": {
"url": "https://bugezy.dev/mcp"
}
}
}
Features
- Voice Input — Chinese, Cantonese, English (Japanese, Korean, Vietnamese coming soon)
- Dual Voice Engine — Web Speech API (free) + Groq Whisper (paid, high accuracy)
- 6 Recording Modes — Record, Rewind 30s, Screenshot, Keyboard, Monitor, CLI
- Bug Capture 10/10 — Console, Network, Resource errors, Web Vitals, DOM replay, Storage, Voice, Screenshots
- Privacy First — Sensitive data auto-masked (PII, JWT, API keys, credit cards)
- AI Auto-correction — Voice transcript cleanup + summarization
- Save 93% Tokens — Structured MCP data vs raw screenshots
Python / Node CLI
npm install -g bugezy-watch
Captures Python tracebacks, Node.js errors, environment snapshots, and PII-masked terminal logs.
Pricing
| Plan | Price | Includes |
|---|---|---|
| Free | $0 | 10 recordings, 5 rewinds, 20 MCP calls/month |
| Monthly | NT$80/mo (~$2.50) | Unlimited everything |
| Day Pass | NT$20 (~$0.65) | 24-hour full access |
Security
- Fable5 4-round audit: 9.5+/10
- Supabase RLS on all 6 tables
- CSP with frame-ancestors
- Session token fragment-based (never in URL query string)
- ECPay payment with idempotent callback
Tech Stack
Chrome Extension (TypeScript) · Cloudflare Workers · Supabase · R2 · Groq Whisper · ECPay
Links
- 🌐 Website: bugezy.dev
- 🧩 Chrome Web Store: Install
- 📖 AI Guide: SKILL.md
- 📋 Features: bugezy.dev/features
- 📝 Changelog: bugezy.dev/changelog
- ❓ FAQ: bugezy.dev/faq
License
Proprietary — © 2026 BugEzy
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。