HELP CENTRE

How to use Csvshape & get help

Everything you need to use the CSV column profiler, understand its limits, and reach a human when something is wrong.

GETTING STARTED

Four steps to a column profile

No sign-up, no account, no setup. Open the app page and you are ready.

  1. 01

    Paste or upload

    Paste CSV text directly into the input area, or click Upload CSV to select a file from disk. Files up to approximately 100 MB are supported on the free tier — if your file is larger, the tool shows a clear error rather than truncating silently.

  2. 02

    Analyze

    Press Ctrl+Enter or click Analyze. The parser detects the delimiter (comma, tab, semicolon) automatically, identifies the header row, and profiles every column.

  3. 03

    Read the profile

    A summary card shows row and column counts. Below it, the column profile table lists each column with detected type, match percentage, blank count, and mismatch count. Click any mismatch count to see the specific offending rows and values.

  4. 04

    Export

    Click Download profile CSV to save the full profile as a CSV file. One click, no sign-in, no data leaves your machine.

FREQUENTLY ASKED QUESTIONS

Common questions about Csvshape

What kind of files can I analyze?
Csvshape works with CSV (comma-separated), TSV (tab-separated), and semicolon-delimited text files. The delimiter is detected automatically by sampling the header row. Files can be pasted as text or uploaded from disk. The free tier supports files up to approximately 100 MB — the limit is your browser's available memory, not an arbitrary cap on the server (because there is no server).
Does the tool send my data anywhere? Is anything stored?
No. The entire analysis runs in your browser using JavaScript. No data — not a single byte — is transmitted to any server, API, or third party. When you close the tab, the profile is gone. The only thing stored between sessions is your theme preference (light/dark mode), saved in the browser's localStorage. No CSV data, column profiles, or personal information is stored anywhere.
What does a "mismatch" mean?
A mismatch is a value that does not fit the dominant type of its column. For example, if a column contains mostly numbers but one cell reads N/A, that cell is flagged as a mismatch. Csvshape lists the row number, the value, and the type the value actually looks like. This catches things like date format inconsistencies, stray text in numeric fields, and mixed delimiter use that a spreadsheet would never flag.
How accurate is the type detection?
The type detection is probabilistic, not guaranteed. It works well on well-structured data: columns where most values share a single format. Edge cases — ambiguous date formats (MM/DD vs DD/MM), mixed types, very short columns, or values that happen to look like a different type — are reported as mismatches rather than silently misclassified. Always verify critical findings against the raw data. The detection is a diagnostic aid, not a certification.
Can I save or share my results?
Yes. Click Download profile CSV to export the full profile as a CSV file. The export includes every column's detected type, match percentage, blank count, and the specific mismatched values. You can attach this file to a bug ticket, include it in documentation, or keep it for your records. There is no account, no cloud storage, and no shareable link — profiles exist only on your machine.
Is there a paid plan?
No. Csvshape is free for everyone. There are no accounts, no subscriptions, and no paid tier. Every feature — type detection, blank counts, mismatch analysis, row validation, export — works for every user on every visit. The only limit is what your browser can hold in memory, and that limit applies to everyone equally.

KNOWN LIMITATIONS

What Csvshape does not do

CONTACT

How to get help

No contact method has been set up for this product yet. A support address and response times will be published here once the operator establishes them.

The operator of this tool is Agen2ic, LLC, a Colorado limited liability company.