Local Multi-File Secret-Pattern Scanner

Scan local UTF-8 files or pasted text for supported secret shapes without exporting matches.

On-device No sign-up Free to use

Scan selected files controls

?

Used in pasted-text mode. Up to 1 MiB of UTF-8. Matches and snippets are omitted from the report.

Drop your CSV files here

or choose them from your device

Up to 100 UTF-8 text files, 1 MiB each and 10 MiB in total. Archives, images, spreadsheets and other binary files are refused rather than guessed at.

    ?

    Optional, up to 50 literal terms of 200 characters each. Matching is literal and a term is never written into the result.

    Processed in your browser. Your inputs stay on this device.

    Showing an example. Change the fields, then use scan selected files to update it.

    Result

    Showing an example.

    Find an example AWS access-key ID and a PEM private-key header on two separate lines.

    2 matches across 1 of 1 file.
    synthetic.txt | PEM private key block | 3
    synthetic.txt | AWS access key ID | 2

    Loading the tool. Your result will appear here.

    https://usefreetools.net/local-multi-file-secret-pattern-scanner · 20 November 2026

    Was this helpful?

    How to use Local Multi-File Secret-Pattern Scanner

    1. Choose Local text files and select supported UTF-8 files, or choose Pasted text for one text sample. Archives and binary formats are outside this scan.
    2. Select the documented credential shapes to look for. Add any literal terms one per line, or remove the example internal-only term if it is irrelevant.
    3. Run Scan selected files. Use file labels, line numbers and pattern names to inspect the original text yourself; the report omits matched values and snippets.
    4. Download the findings CSV if needed. It includes collected findings beyond the 200 on-page rows, but match counts stop at 2,000 per pattern or literal term per file.

    Example: Local Multi-File Secret-Pattern Scanner

    Find an example AWS access-key ID and a PEM private-key header on two separate lines.

    You add
    Look for PEM private key blocks: true Look for AWS access key IDs: true Look for GitHub tokens: true Look for Slack tokens: true Extra literal terms, one per line: internal-only Text files to scan: synthetic.txt contains "Public demo\nAKIAABCDEFGHIJKLMNOP\n-----BEGIN PRIVATE KEY-----\n"
    You get
    2 matches across 1 of 1 file. synthetic.txt | PEM private key block | 3 synthetic.txt | AWS access key ID | 2

    Options

    Input source
    Choose Local text files, Pasted text.
    Text to scan locally
    Used in pasted-text mode. Up to 1 MiB of UTF-8. Matches and snippets are omitted from the report.
    Stripe-shaped secret key
    Looks for the supported sk_ or rk_ live/test key shapes. A match reports the file, line and pattern name without printing the key.
    Google API key identifier
    Looks for the supported AIza-prefixed identifier shape. A match does not establish that the key is active or grants access.
    OpenAI-shaped key
    Looks for the supported sk- key shapes, including the listed project and service-account prefixes. Formats outside those patterns can be missed.
    Twilio API key SID
    Looks for SK followed by 32 hexadecimal characters. This identifies a SID-shaped value, not its associated secret or permissions.
    Text files to scan
    Up to 100 UTF-8 text files, 1 MiB each and 10 MiB in total. Archives, images, spreadsheets and other binary files are refused rather than guessed at.
    Look for PEM private key blocks
    Flags a supported PRIVATE KEY header line. The report omits the key text and does not validate or decrypt the block.
    Look for AWS access key IDs
    Flags supported AKIA or ASIA access-key ID shapes. It does not check the corresponding secret, account or live permissions.
    Look for GitHub tokens
    Flags the supported GitHub token prefixes and lengths. New or unlisted token formats can be missed.
    Look for Slack tokens
    Flags the supported xox token shapes. It does not contact Slack or test whether a token still works.
    Extra literal terms, one per line
    Optional, up to 50 literal terms of 200 characters each. Matching is literal and a term is never written into the result.

    Supported inputs and limits

    • UTF-8 text only: up to 100 files, 1 MiB each and 10 MiB combined, or one pasted input up to 1 MiB.
    • Up to 50 extra literal terms of 200 characters each.
    • Counts stop at 2,000 per enabled pattern or literal term per file; they are capped findings, not an exhaustive count.
    • On-page rows show at most 200 findings;
    • CSV includes all collected rows.
    More details

    Matched values and snippets are omitted, but filenames and line locations are reported.

    No archive/OCR scan, breach search, live token check or complete detection is provided.

    Where your input is processed

    This tool does not upload your inputs. See the limits for anything saved on your device.

    A finding is a review location

    The report identifies a file, line and pattern class without displaying the matched value or surrounding source. Fixed patterns can produce false positives, while unlisted or altered secrets can be missed. Review findings in your own files and do not treat a clear report as proof that a collection contains no secrets.

    Questions about Local Multi-File Secret-Pattern Scanner

    Are secret values printed?

    No. Findings omit values and source snippets.

    Does zero mean no secrets exist?

    No. It means no supported pattern or entered literal term was found in the selected supported text.

    Does it test whether a token works?

    No. It makes no authentication or network requests.

    Page updated · Project managed by Tony Hines · Report a problem with Local Multi-File Secret-Pattern Scanner