Item logo image for Wusel Capture

Wusel Capture

Item media 1 (screenshot) for Wusel Capture

Overview

Screenshot a page (full / viewport / region), annotate it, and send it to Claude Code (or export an annotated image + JSON context).

Wusel Capture turns a design change into a task your AI coding assistant can ship. See something on a web page you want changed? Screenshot it, mark up exactly what should change, add a note, and send it straight to Claude Code or Codex in VS Code. Your assistant gets the annotated image plus the page's technical context, finds the code behind the page, and implements the change — using your existing components, tokens, and patterns. No more describing a bug in words, hunting for the right file, or pasting screenshots by hand. HOW IT WORKS 1. Capture the page you want to change. 2. Annotate what should change, and write a short note for the AI. 3. Send it to Claude Code or Codex in VS Code — or download the image and context. WHAT YOU CAN CAPTURE - The full page, including everything below the fold. - The visible viewport. - A region you draw yourself. ANNOTATE - Text, arrows, and boxes to point at exactly what should change. - A redaction tool that pixelates sensitive areas before you send. CONTEXT, NOT JUST PIXELS Every capture bundles the page URL together with console errors and failed network requests, so your assistant sees what you saw — not just a picture. This is what lets it reproduce the issue and land the fix in the right place. Token-like URL parameters are redacted automatically, and no cookies, headers, or request bodies are ever captured. PRIVATE BY DESIGN Captures are created and processed locally in your browser and go straight to your own assistant. Nothing is sent to us — there are no servers, no accounts, no analytics, and no telemetry. When you send a capture, your own Claude Code or Codex processes it under your own account, exactly as if you had pasted it in yourself. OPEN SOURCE Wusel Capture is free and MIT-licensed. Read exactly what it does, line by line, at github.com/secretlifeof/wusel-capture. GET STARTED Add the skill your assistant uses to understand a capture: npx skills add wusel-capture Then pin the extension and take your first capture. SUPPORTED ASSISTANTS - Claude Code (VS Code) — the default. - Codex (VS Code). OPTIONAL: ZERO-PROMPT DELIVERY Install the optional open-source native host and captures are written straight into the matching local project folder, with no "Open VS Code?" prompt in between.

Details

  • Version
    0.1.0
  • Updated
    July 18, 2026
  • Size
    200KiB
  • Languages
    English
  • Developer
    Email
    wusel-capture@wusel.app
  • Non-trader
    This developer has not identified itself as a trader. For consumers in the European Union, please note that consumer rights do not apply to contracts between you and this developer.

Privacy

Manage extensions and learn how they're being used in your organization

Wusel Capture has disclosed the following information regarding the collection and usage of your data. More detailed information can be found in the developer's privacy policy.

Wusel Capture handles the following:

Web history
User activity
Website content

This developer declares that your data is

  • Not being sold to third parties, outside of the approved use cases
  • Not being used or transferred for purposes that are unrelated to the item's core functionality
  • Not being used or transferred to determine creditworthiness or for lending purposes

Related

Screenshot to Editable Design - Visily

4.4

Turn any website or app screenshot into a fully editable wireframe or UI design. Capture, redesign, and prototype in Visily.

Pictotext — Image Text Extractor

5.0

Extract text from any image on the web or any screen region using local Tesseract.js OCR. No server. No API key. Private by design.

HTML Text to .md — Local Markdown Web Clipper

5.0

Select any part of a page and convert it to clean Markdown. Tables, code and math stay intact. Runs locally in Chrome.

Annotator - for Chrome

5.0

Capture page screenshot, annotate with draws/highlights/arrows/text, select & move elements, and send to IDE.

HAVI – Human-Agent Visual Interface

0.0

Capture visual annotations on any web page; your AI coding agent pulls the full context in via MCP.

DoMarkdown - WebPage, HTML, and Reddit to Markdown Files

0.0

Export web, HTML, and Reddit pages to Markdown files. Turn webpage context into Markdown for Claude Code and Gemini.

Clasp-it

5.0

Pick any webpage element and send its context to your AI editor via MCP

Decant — Send any page to Claude, ChatGPT & Gemini

5.0

Extract any web page & send to Claude, ChatGPT or Gemini in one click. Clean Markdown, JSON, MCP bridge for AI agents & developers.

Full Page Screenshot

4.8

Capture full-page screenshots with one click. Annotate, crop, blur, and export as PNG, JPG, WebP, or PDF. 100% local.

Screen Capture & Editor - Go Full Page

3.0

Screen capture any area or go full page to grab the entire website, then edit and export in seconds.

Full Page Screen Capture and Area Screenshot

5.0

Full page screen capture, visible window or area screenshot. Any webpage as a lossless PNG - copy to clipboard or save as PNG.

Claude Code Context Capturer

0.0

Convert any web page into Claude Code-friendly Markdown context with one click.

Screenshot to Editable Design - Visily

4.4

Turn any website or app screenshot into a fully editable wireframe or UI design. Capture, redesign, and prototype in Visily.

Pictotext — Image Text Extractor

5.0

Extract text from any image on the web or any screen region using local Tesseract.js OCR. No server. No API key. Private by design.

HTML Text to .md — Local Markdown Web Clipper

5.0

Select any part of a page and convert it to clean Markdown. Tables, code and math stay intact. Runs locally in Chrome.

Annotator - for Chrome

5.0

Capture page screenshot, annotate with draws/highlights/arrows/text, select & move elements, and send to IDE.

HAVI – Human-Agent Visual Interface

0.0

Capture visual annotations on any web page; your AI coding agent pulls the full context in via MCP.

DoMarkdown - WebPage, HTML, and Reddit to Markdown Files

0.0

Export web, HTML, and Reddit pages to Markdown files. Turn webpage context into Markdown for Claude Code and Gemini.

Clasp-it

5.0

Pick any webpage element and send its context to your AI editor via MCP

Decant — Send any page to Claude, ChatGPT & Gemini

5.0

Extract any web page & send to Claude, ChatGPT or Gemini in one click. Clean Markdown, JSON, MCP bridge for AI agents & developers.

Google apps