Skip to content
patchkind

Website feedback · Reviewed changes

A website feedback tool where the note comes back as a pull request

A website feedback tool usually gives you a list. Visitors leave notes, sometimes with a screenshot, and you get a dashboard full of things people want changed. The list grows faster than you work through it.

Patchkind closes that gap with you deciding at each step. A visitor leaves a note on your site. Patchkind turns it into a short plan. You read the plan, edit it, approve it. Nothing runs until you press Run. Then Claude Code works inside your own GitHub Actions, on your own Claude subscription token or API key, and opens a pull request in your repository. You review the PR. Patchkind never merges.

How a note becomes a PR

  1. Note. A visitor clicks the feedback button on your site and says what is confusing or missing. A screenshot is optional. Visitors do not need an account.
  2. Plan. Patchkind drafts a short plan. You edit it and approve it. Approving does not start anything.
  3. Run. You press Run. Claude Code runs in your repository's GitHub Actions. Your Claude credential lives in your GitHub secrets; Patchkind does not hold it.
  4. Review. A pull request comes back with the checks that ran. You read it and decide whether to merge. Claude may ask a question or stop instead of writing code.

A real one: explain what a company does

On AlgoThesis, a stock-news site run by Patchkind’s owner, the documented scope of PR #720 was to add one sentence under the company header using the site’s existing sector and industry data. This summarizes the PR record, not the original reporter note.

The plan was approved and run. The pull request changed 1 file, +13 lines. The owner merged it on 5 October 2026. Review time was not measured.

A follow-up improved layout and naming: the sentence on its own line, company names displayed differently. 2 files, +18 / −7. Also merged. Merged is not the same as done, and the tool is built for that loop.

Live on algothesis.aiNVIDIA Corp
NVIDIA Corp operates in the Semiconductors & Related Devices industry.
Text from the live header after #720 and the layout follow-up #723. View header crop ↗

Read the AlgoThesis case study for PRs #720 and #723 and what the trial does not prove.

Who it is for

  • Websites and web apps built with npm, in private or internal GitHub repositories, running GitHub-hosted Actions runners.
  • Solo builders and small teams who get good suggestions and never get to them.
  • Owners who want to read every change before it goes live.

Who it is not for

  • Native mobile apps. There is no iOS or Android SDK; a mobile app can open the hosted feedback link instead.
  • Surveys, NPS or CSAT. Patchkind collects notes and screenshots, not scores.
  • Sites with no code to change, such as a hosted page builder with no GitHub repository.
  • Teams who want changes merged without a person reading them. Patchkind never merges.
  • Projects on other stacks or other AI providers. The beta supports npm projects and Claude Code. Dependency changes are not allowed in a PR.

How it compares to BugHerd, Marker.io and Userback

BugHerd, Marker.io and Userback are mature website feedback tools. They support more website stacks than Patchkind’s current npm coding workflow, they have been around longer, and they connect to more issue trackers and project boards. Their core reporting flows organize tickets and annotations. Marker.io also supports AI-assisted fixes through MCP.

Patchkind’s coding workflow aims to return a pull request you review. Claude may instead ask a question or stop. It gets there by having you approve a plan first and by running the coding step in your own GitHub Actions. It is in a free invited beta, it supports npm web apps and Claude Code, and our published trial is on one of our owner’s own products. If you need broad stack support and a polished triage workflow today, those tools are the safer pick. If you want visitor notes to come back as small, reviewable code changes, that is what Patchkind does.

Marker.io’s MCP integration lets connected agents read report context and open PRs, with autonomy chosen by the owner. Userback also connects feedback to coding assistants. Compare the plan and approval workflow, not a ticket-versus-code divide. Patchkind vs Marker.io →

Frequently asked questions

Can visitors suggest design changes or missing features? Yes. Feedback can describe a confusing layout, missing information or an idea, as well as something broken. You decide which notes deserve a plan; the beta is for small, bounded changes.

Do I have to show a floating button? No. You can share the hosted feedback link without adding code. With the widget script installed, a custom footer or menu link can open the same panel instead of the floating launcher. See the feedback widget.

Who decides when a change goes live? You approve the plan, separately press Run, then review any resulting pull request. Patchkind does not merge or deploy it. In our trial, the owner closed a PR whose plan pointed at the wrong route; review is still part of the process.

Do my visitors need to sign in? No. Visitors leave a note without an account. You, the owner, sign in with ChatGPT. Optional AI follow-up questions and feedback summaries use OpenAI and can send feedback text and screenshots to that provider; see the privacy notice. That is separate from the Claude coding run in your GitHub.

What does it cost? The invited beta is free. Your own Claude subscription or API charges and GitHub Actions usage limits still apply.

Try it on your site

Free invited beta. npm projects, GitHub-hosted runners, Claude Code.

Request free beta access

Free invited beta. This requests an invitation, not instant access.

Questions: support@patchkind.com

See the Claude Code workflow guide and hostile-feedback test.

Read the AlgoThesis case study →Explore the feedback widget →