Web Tables → Excel — verified table export

    1

    Export the tables on any web page into a real .xlsx workbook (or CSV), then read the file back to prove it is correct. Handles merged cells, admin-panel grids and code-like values such as 00123.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Web Tables → Excel — verified table export

    Example session with this skill installed

    Export the population table on https://en.wikipedia.org/wiki/List_of_countries_and_dependencies_by_population to Excel, and tell me exactly what ended up in the file.

    • Read your context and instructions
    • Compiled the web tables excel
    • Generated the data export

    Summary

    The page has 1 data table. I exported it to population.xlsx (14.7 KB) and read the file back to check it.

    SheetRowsColumnsNotes
    List of countries and territories by total population240 data rows + 1 header7Rank column rebuilt (the page draws it with CSS)
    Report148Source URL, page title, timestamp, table sizes, known gaps

    What is in the file (first rows, read back from the .xlsx)

    #LocationPopulation% of worldDateSourceNotes
    World8,232,000,000100%13 Jun 2025UN projection[1][3]
    1India1,429,404,00017.3%1 Jul 2026Official projection[4][b]
    2China1,404,890,00017.0%31 Dec 2025Official estimate[5][c]

    Checks I ran

    1. Sheet count = 1 table + Report. ✔
    2. First data rows match the page (World, India, China). ✔
    3. Rank column: the page shows ranks via CSS, not text; they were rebuilt as 1..N. The "World" total row has no rank on the page, so it stays blank. ✔
    4. Row count plausible: 240 rows (World + countries and dependencies). ✔

    Known gaps

    • Population figures keep the page's thousands separators, so they are stored as text. Convert the column with Excel's Text to Columns if you need to calculate with it.
    • Footnote markers such as [1][3] are kept as they appear on the page.

    Next: tell me if you want CSV instead, or another table from the same page.

    population.xlsx

    CSV · data export

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Most "table exporters" hand you a file and hope. This one reads the workbook back off disk and checks it before you ever see it.

    What it does

    • Finds every table on the page you point it at: real <table> markup, ARIA data grids, admin-panel grids (Element UI / Element Plus split headers, ag-Grid pinned columns, MUI DataGrid, react-data-grid) and div-based grids as a best effort. Each one is listed with row count, column count, header detection and warnings before anything is exported.
    • One sheet per table plus a Report sheet with the source URL, page title, timestamp, per-table sizes and every known gap.
    • Merged cells handled honestly: a rowspan/colspan group keeps its text in the anchor cell and leaves covered cells blank, so rows stay aligned with the page. Multi-row group headers are kept, not flattened.
    • Codes stay codes: 00123, SO-0001, 19-digit order numbers and phone numbers stay text, so Excel cannot turn them into 123 or 1.23E+18. Clean numbers are written as numbers, so you can sum them.
    • Row numbers drawn by CSS are rebuilt instead of silently shifting every column.
    • Verified: a built-in inspector re-opens the .xlsx and prints the real cell contents, so "the export worked" is a claim with evidence.
    • CSV output when another tool needs it.

    Good for

    Quotations, price lists, admin-panel order lists, report tables, leaderboards, listings and dashboards that only exist as HTML.

    Requirements

    Node.js 22+ and an installed Chrome, Edge or Chromium. No npm packages: three self-contained scripts using only Node built-ins. The page is opened in a throwaway private browser profile; nothing is sent anywhere except to the page you asked for.

    Known limits, stated up front

    • Only content already loaded in the page is captured; paginated or virtualised rows that were never rendered are absent (use the scroll option and check counts).
    • Cross-origin iframes cannot be read.
    • Rich formatting, images and cell colours are not preserved; no Excel merge ranges are written.
    • One page per run. Content behind a login is only read when you start your own browser with remote debugging and ask for it; nothing is bypassed.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Data & Databases.

    Node 22+ required. Browser must be local. Only captures DOM-resident content. No pagination clicking. No images or cell styling. Cross-origin iframes unsupported.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 14 days ago

    • Free with an Agensi account

    Software
    Node.js 22+
    Google Chrome
    orMicrosoft Edge
    orChromium

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Browser
    Network Access

    Terminal: runs the three Node scripts. Browser + network: starts a local Chrome/Edge in a throwaway profile and loads only the page you asked for. Read/Write files: saves the data file and the workbook in your working folder. No environment variables, no credentials, no telemetry.

    Listed14 days ago

    What's inside

    Frequently Asked Questions

    This skill handles complex structures like ag-Grid and MUI DataGrid that standard scrapers often miss. It specifically prevents Excel from mangling data by preserving leading zeros and long numeric IDs as text, ensuring high-fidelity data transfers.

    It captures every table row that is currently rendered in the page. Lazy-loaded rows are included when you use the scroll option (`--scroll N`), which scrolls to the bottom N times first. It does not click 'Next' on paginated results or scroll inside virtualised grids, so for those, open the view you want and compare the row count with the page.

    Three self-contained Node.js scripts (extract.mjs: page to data file; to-xlsx.mjs: data file to .xlsx or CSV; inspect.mjs: reads a workbook back), the SKILL.md workflow your agent follows, and a README. No npm packages and no configuration files.

    This skill is designed for Node.js 22+ environments and works best with agents that have local file system access and can execute JavaScript. It requires a Chromium-based browser (Chrome or Edge) installed on the host machine.

    inspect.mjs re-opens the written .xlsx and prints the real cell contents of every sheet (or all cells as JSON). The SKILL.md workflow makes your agent check sheet count, the first rows against the page, codes kept as text, merged headers and row counts before it reports success. Every workbook also carries a Report sheet with the source URL, timestamp, table sizes and known gaps.

    New in Data & Databases

    Free