UsrHelper

See It in Action

A bug report in two clicks: take the shot, mark what is wrong, and the report leaves with the page address, the real browser and system, the last clicks and the errors the page printed — the details a developer would otherwise have to ask for.

Rather watch it on YouTube? Open the demo in a new tab.

Project Overview

A Chrome extension that makes software-deployment feedback effortless. Testers and end users capture annotated screenshots or record screencasts with voice narration, and the report is saved locally with a timestamp and/or handed off to email — no backend, no accounts, no tracking.

The Problem It Solves

During software rollouts, the feedback loop is usually the weakest link: users describe bugs in vague emails, screenshots arrive uncropped and full of personal data, and “it doesn’t work” tickets bounce back and forth for days. Dedicated feedback platforms fix this but demand accounts, cloud storage, and a per-seat subscription — overkill for a deployment team that just needs clear, annotated evidence attached to an email.

Capture: Three Ways to Take the Shot

The popup is the whole interface. Pick a capture mode and the editor opens on the result; recent reports stay in a list underneath, with a thumbnail, “Show file” and “Email again”.

• Visible area — what is on screen right now.
• Full page — the extension scrolls and stitches the whole document automatically.
• Selected region — drag a rectangle around the part that matters.
• A timestamp lands in the file name and in the corner of the image, so an old screenshot can never be passed off as today’s.

The popup: three capture modes, screencast recording, and a report history with thumbnails and one-click re-email.

Annotate, Explain, Anonymise

The editor is a hand-rolled <canvas> object layer — every mark stays a movable object until you save, so nothing is destructive except the one thing that should be.

• Freehand marker, rectangle, ellipse, arrow, text, numbered step markers, crop, full undo/redo.
• Yellow sticky notes for a sentence or two of context: numbered on the image, sized to stay readable from a small region crop to a retina capture, and repeated as text in the companion .json so a developer can copy them instead of retyping.
• Anonymisation brush — paint a pixel mosaic over customer names, invoice numbers or tokens. It is baked into the exported image, so what was hidden is gone for good.

A real annotated capture: numbered sticky notes, arrows, freehand marks — and one row blurred out with the mosaic brush.

Click Path: What the User Actually Did

“I clicked something and it broke” is the hardest ticket to reproduce. One button stamps the recent clicks onto the screenshot as numbered markers, and the same sequence — with coordinates, scroll position and exact times — goes into the report file.

“Add click path” turns the last clicks into numbered markers on the image; below, the screencast panel with the microphone meter and recording limits.

Screencasts With Narration

When a still cannot show it, record it. Pick a tab, a window or the whole screen, talk over it, and the file lands in the same folder as the screenshots.

• Microphone level meter, 3-2-1 countdown, pause/resume and an elapsed timer.
• Optional camera bubble and a clock burned into the video.
• Clicks appear as ripples and keyboard shortcuts as captions — while ordinary typing is never shown, so a password typed during a recording does not end up on screen.
• Automatic clip splitting: long sessions are cut into standalone clips (default 5 minutes each, 30 minutes total, configurable) instead of one unsendable attachment.
• Frames are encoded in a worker rather than in the page, so switching tabs mid-recording does not drop the picture.

The Report a Developer Can Act On

Every capture is saved to a folder of your choice inside Downloads/, with the date and time in the name, next to a companion .json that carries everything a first reply would otherwise have to ask for.

• Your description and your sticky notes, as text.
• The page address and title, the exact capture time, and the click path.
• JavaScript errors the page printed — scoped to the project domains you configure.
• The real environment: operating system and version, processor architecture, exact browser build, screen size and device pixel ratio. Browsers understate this on purpose — every Apple Silicon Mac reports itself as an Intel one — so UsrHelper reads the accurate values instead.
• One click opens a ready-made email with recipients, subject and description already filled in, and the extension reminds you which file to attach.

The companion .json: reporter, capture time, page, environment, console errors, click path, sticky notes and the files that belong to the report.

Project Profiles You Can Hand to Testers

One tester works on three systems, each with its own recipients and its own rules. Profiles keep them apart, and a profile can be shared as a file so nobody has to type an address by hand.

• Per profile: email recipients (To/CC), download subfolder, subject prefix, description template, recording limits, and the project domains whose console errors are captured.
• Export the active profile to a file and send it to your testers; importing shows the recipients for confirmation before anything is stored.
• The file carries project settings only — never reporter details, never report history.
• The interface is available in English and Polish.

Exporting a project profile — one file, handed to the whole test team.

Privacy by Design

Everything stays on the user’s machine: captures land in the local Downloads/ folder, settings live in the Chrome profile, and there is no backend, tracking, or analytics of any kind. Screens from software under test routinely show customer data, so nothing is uploaded — which also means the extension works offline and keeps the Chrome Web Store review surface small. The source is MIT-licensed and public, so the claim can be checked rather than taken on trust.

Documentation

UsrHelper ships with a full user guide in English and Polish: browse the online manual with a language toggle, read the Markdown guides with screenshots directly in the repository, or grab the printable PDF manuals (EN + PL) attached to every release.