Show HN: An AI-Powered Widget for Collecting User Feedback

Make This Better launched an AI-powered feedback widget that captures screenshots, DOM snapshots, and console logs when users report issues, then uses AI to clarify and extract the underlying need, and integrates with coding agents like Claude Code, Cursor, Windsurf, and Codex via MCP and CLI to turn feedback into shipped fixes. The tool is free to start with no credit card required, and it aims to close the gap between user feedback and developer understanding, as demonstrated by 23 users clicking Export and leaving without explanation.

Every user signal becomes an issue your agent can ship. Mark-up mode activates. They click an element, draw a circle, write a note. A screenshot, DOM snapshot, and console log are auto-captured in the background. AI asks one or two clarifying questions, then extracts: the scene context , the blocker, and the real underlying need. Install the Skill and connect MCP once left . The Skill gives your agent the workflow; MCP and the CLI provide feedback data and status actions. Then run /makethisbetter pick to inspect the original evidence, claim the feedback, and start the fix right . Works with Claude Code, Cursor, Windsurf, and Codex. The public Feedback Board shows progress in real time. Your user watches their issue move from received → in progress → shipped. Trust earned, churn prevented. AI coding makes you ship fast — not perfect. The real gap isn't between code and deploy. It's between your users and your understanding of what they actually need. Here's how Make This Better closes it: Twenty-three users kept clicking Export, then left. Analytics and logs looked normal. The widget lets them report the problem on the spot, and your agent takes over. A user says "it doesn't work," and you have no idea why. The widget captures the URL, errors, and screenshot. Run /feedback pick , and your agent gets to work. Without a QA team, you click around before launch and hope for the best. Real users still hit problems you never imagined. The widget captures what happened, turning every session into a test. Seven users ask for email sharing. AI turns their scattered feedback into one issue, and your agent builds it. A paid feature you nearly missed is now obvious. Install the CLI, connect MCP, and start picking up feedback — all from your terminal. Your users give feedback. Your agent ships the fix. You sleep. Need the widget for your users? Docs /widget Pay for your team. Bring your own API key for free AI on any plan. Stop guessing what’s broken. Let your users tell you — and let your agent ship the fix. Free to get started. No credit card required.