From plugin-windows-mcp
This skill should be used when the user asks to "scrape a webpage", "extract web content", "read the clipboard", "write to clipboard", "copy and paste", "select multiple files", "batch select items", "fill multiple form fields", "multi-edit", or needs to extract data from web pages, manage clipboard content, or perform bulk selection and input operations on Windows.
How this skill is triggered — by the user, by Claude, or both
Slash command
/plugin-windows-mcp:web-and-data-operationsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
This skill covers four data-oriented tools: Scrape (web content extraction), Clipboard (read/write clipboard), MultiSelect (batch item selection), and MultiEdit (bulk form input).
This skill covers four data-oriented tools: Scrape (web content extraction), Clipboard (read/write clipboard), MultiSelect (batch item selection), and MultiEdit (bulk form input).
Extract information from web pages by URL.
Key parameter: url (required — the page to scrape)
Returns: Extracted page content as text.
| Scenario | Use Scrape | Use Snapshot (DOM) |
|---|---|---|
| Page is already open in browser | No | Yes |
| Need data from a URL not yet visited | Yes | No |
| Need to interact with page elements | No | Yes |
| Just need the text content | Yes | Either |
| Need to fill forms or click links | No | Yes |
Basic content extraction:
Scrape(url="https://example.com/data-page")
-> process extracted text
Sequential multi-page extraction:
Scrape(url="https://example.com/page/1") -> extract data
Scrape(url="https://example.com/page/2") -> extract data
-> combine results
Best practices:
Clipboard(write) to store extracted dataSnapshot with use_dom=True)Read from or write to the Windows system clipboard.
Clipboard(action="read")
Returns the current clipboard text content.
Clipboard(action="write", text="Data to paste")
Writes text to the clipboard for pasting.
Extract selected text from any app:
Shortcut("ctrl+a") -> select all
Shortcut("ctrl+c") -> copy to clipboard
Clipboard(read) -> retrieve the text
Paste specific content into any app:
Clipboard(write, text="Prepared content")
Click(target_field_x, target_field_y)
Shortcut("ctrl+v") -> paste
Transfer data between applications:
App(switch to source) -> Wait(0.5)
Shortcut("ctrl+a") -> Shortcut("ctrl+c")
text = Clipboard(read)
-> process or transform the text
Clipboard(write, text=processed_text)
App(switch to destination) -> Wait(0.5)
Click(target_field) -> Shortcut("ctrl+v")
Best practices:
ctrl+c — other actions may overwrite itSelect multiple items simultaneously (files, folders, checkboxes, list items).
Key parameter: items (array — list of items to select, by coordinates or identifiers)
Select multiple files in File Explorer:
Snapshot -> identify file positions
MultiSelect(items=[file1_coords, file2_coords, file3_coords])
Select checkboxes in a form:
Snapshot -> identify checkbox positions
MultiSelect(items=[checkbox1, checkbox2, checkbox3])
Advantages over individual clicks:
Click operationsBest practices:
Snapshot first to identify item positionsScreenshot after MultiSelectInput text into multiple form fields simultaneously.
Key parameter: fields (array — field identifiers with their values)
Fill a complete form:
Snapshot -> identify all form fields
MultiEdit(fields=[
{field: field1_id, value: "First Name"},
{field: field2_id, value: "Last Name"},
{field: field3_id, value: "email@example.com"}
])
Advantages over sequential Type:
Best practices:
Snapshot first to identify field IDsScreenshot after MultiEditClick and GUI interaction insteadScrape(url) -> extract data
-> format as tab-separated values
Clipboard(write, text=formatted_data)
App(switch to Excel) -> Wait(0.5)
Click(target_cell) -> Shortcut("ctrl+v")
App(launch "File Explorer") -> Wait(2)
Navigate to folder via address bar
Snapshot -> identify files
MultiSelect(items=[target_files])
Shortcut("ctrl+c") -> copy files
Navigate to destination folder
Shortcut("ctrl+v") -> paste files
App(switch to browser) -> Wait(0.5)
Snapshot -> identify all form fields
MultiEdit(fields=[all_field_values])
Snapshot -> identify submit button
Click(submit_x, submit_y)
Wait(1) -> Screenshot -> verify submission
ctrl+c completed before reading. Add Wait(0.3) between copy and readSnapshot to verify field typesFor advanced data extraction and manipulation patterns:
references/data-patterns.md — Advanced scraping strategies, clipboard automation chains, and bulk operation patternsnpx claudepluginhub mustafaakben/plugin-windows-mcp --plugin plugin-windows-mcpControls desktop GUI as a fallback when APIs, CLIs, file editing, and browser automation are unavailable or have failed. Clicks, types, reads screen, and drives native apps on Windows/macOS/Linux.
Drive a real browser with agent-web-interface MCP tools (navigate, snapshot, click, type, screenshot). Useful for automating live web interactions, extracting selectors, and validating page state.
Automates browser tasks like form filling, data extraction, and multi-step web workflows using Yutori Navigator agent. Useful for website interactions requiring clicking, typing, or navigation.