NorthstarDashboardProjectsTeamAN

Project overview

Launch readiness

Everything your team needs before the next release.

Open tasks12

Resolved84%

Selected elementPublicar proxima version

Feedback

Report feedbackPoint to what needs attention.

BugTranslationIdea

What happened?This should say “Publish next version” in Spanish.

Button selectedScreenshot masked

Send feedback

Page, locale, browser and selected element attached

Feedback

Let users show you what needs fixing.

An embeddable feedback widget for bugs, AI slop, bad translations, confusing details, and ideas—captured with the page context your team or coding agent needs to act.

Add Feedback Try the live demo

One script · Isolated from your app’s styles · Free during beta

Catch it where it happens

“Something is wrong” is not enough.

A screenshot in chat still leaves someone asking which page, which element, which browser, and what the user expected. Feedback starts inside the product, while the useful context is still available.

The user stays in the product

They point at the relevant element and explain the problem in the moment, without opening a support portal.

The context stays with the report

Page, environment, selected element, screenshot, and the user’s own words arrive together.

The next builder can act

Your team or coding agent starts with the report instead of reconstructing it.

Made for what slips through

The messy edges of software that keeps changing.

AI slop

Awkward generated copy, missing states, inconsistent UI, and details that only looked acceptable in a prompt.

Bugs

Let users point to the exact element and explain what happened without opening another tool.

Bad translations

Report the wrong phrase where it appears, with the page and locale context intact.

Ideas & confusion

Capture suggestions and unclear behavior at the moment a user notices them.

More than a message

Capture the moment, not a reconstruction of it.

The reporter can select the relevant element, describe the problem, and include an annotated screenshot. iterateTo keeps the report beside the technical context needed to investigate it.

Page and selected element
Confidence-ranked selector candidates
Browser, operating system, and viewport
Optional annotated screenshot
Privacy masking and stripped URL queries
The user’s own words

Report to action

Work on feedback without losing the details.

01
Point and report
The user shows you where the problem is and explains what they expected.
02
Triage together
Review the raw report, priority, tags, assignee, related feedback, notes, replies, and activity.
03
Hand off complete context
Copy a coding-agent prompt or let agents and automations work the queue through MCP and REST.

Built to live inside real products

Easy to add. Careful about what it touches.

One script adds the widget to a website or web app. Its interface is isolated from the host page, capture code loads only when needed, and project settings control appearance and allowed origins.

<script src="/w/v1.js" data-key="pk_..." />

The same queue, available to agents

Do not make the coding agent reconstruct it either.

Use copy-ready prompts, REST, or MCP to give an agent the original report and its technical context.

feedback_get({
  id: "fdb_01JY...",
  include: "full_context"
})

→ selected element
→ browser + viewport
→ masked screenshot
→ original report

Agent-ready, human-readable

Complete context, with people still in the loop.

Original report

Keep the user’s words, selected element, capture, and environment together.

MCP and REST

Let an agent read and work the same queue your team sees.

Visible changes

Updates, notes, and agent activity remain inspectable by people.

Add it to what you are shipping

Make the next report useful.

Catch the bugs, AI slop, bad translations, confusing details, and ideas that would otherwise be lost.