BugEzy is a Chrome extension that lets developers capture bugs by voice + recording. AI reads the report automatically via MCP and suggests fixes — saving 95% of debugging communication time.
Built for Asian developers: Chinese / Cantonese / Japanese / Korean voice support, an affordable NT$80/mo plan, and MCP integration so AI reads reports directly. Signature features: live monitor, 30-second rewind, Whisper precise voice, terminal CLI, and token transparency.
Any MCP-capable AI tool works, including Claude Desktop, Claude Code, Cursor, VS Code, Zed, Windsurf, Google Antigravity, Gemini CLI, and more. Just one URL:https://bugezy.dev/mcp
BugEzy records DOM structure changes, not screen video. Password fields (type="password") are automatically masked by rrweb, so actual passwords are never captured.
Each report has a random encrypted ID (UUID) that cannot be guessed or searched — only people with the report link can view its content. If you share the link with colleagues or AI tools, they can access the report; unshared report links never appear in any public listing. Tip: avoid posting report links in public places (issues, forums) to prevent unintended access.
Reports are stored on Cloudflare R2 (global CDN); user data on Supabase (PostgreSQL). All transfers are encrypted over HTTPS.
The free plan includes 10 recordings, 5 rewinds, and 20 MCP AI reads per month. Screenshot annotation and live monitor are unlimited. Reports are kept for 7 days.
NT$80/mo (about US$3) unlocks unlimited use of all features, 90-day report retention, plus advanced features like terminal CLI and Whisper precise voice.
Click "Upgrade" in the BugEzy popup and pay by credit card or ATM.
Yes, anytime. After cancelling you keep premium features for the rest of the billing period, then revert to the free plan.
Chrome and all Chromium-based browsers (Edge, Brave, Arc, etc.).
Minimal. BugEzy only records DOM changes while you are actively recording; live monitor mode only captures Console and Network errors, not the DOM.
Model Context Protocol — an open standard from Anthropic that lets AI tools connect to external services. BugEzy MCP lets AI read your bug reports directly, with no copy-paste.
Tokens are the unit AI uses to process text — effectively your AI cost. BugEzy sends reports as structured text (not screenshots), so the same bug info takes 1/20 the tokens. Every MCP AI read shows a token estimate so you can see the savings.