A signal is only
the beginning.

00. Shared access, real-world limits

Profile Finder is a browser interface for quick OSINT checks without installing or configuring the underlying scanner. Requests use shared public server infrastructure. Target services may respond differently because of IP reputation, rate limits, bot protection, regional rules, CAPTCHAs or temporary blocking.

01. From query to result

Username and email inputs use separate validation and classification rules, but they follow the same six-stage research path.

1. 01 / INPUT

   One identifier

   Start with an exact handle or a complete email address.

2. 02 / VALIDATE

   Check the format

   Trim the input and apply the rules for the selected workflow.

3. 03 / REQUEST

   Check each source group

   Request one category and page at a time within each lane.

4. 04 / RECEIVE

   Keep every response

   Retain positive, negative, unknown and error states.

5. 05 / ORGANIZE

   Surface possible matches

   Group positive signals and the details returned with them.

6. 06 / REVIEW

   Return to the source

   Look for independent context before drawing a conclusion.

The interface is an evidence-organizing layer over returned API data. It does not replace the source and does not turn a technical status into an identity finding.

02. Result status reference

The possible-match area is intentionally narrower than the complete response log. The table below describes how the current browser code treats common response states.

Status text, reason fields, and extra fields originate in the scan response. Profile Finder does not silently convert a failed request into a negative result.

03. Username search method

The browser trims the submitted value, rejects an empty username, rejects whitespace, and limits the value to 100 characters. The interface asks for the exact handle without the @ symbol.

Source coverage

What becomes a possible match

Only a row whose status normalizes exactly to Found becomes a possible-account row. Rows are de-duplicated by normalized site name and URL. All other returned rows remain available by category in the response log.

When a valid result URL is present, the interface can display a Google-provided favicon and an “Open source” link. The icon and link improve navigation; neither adds evidentiary weight to the result.

04. Email search method

The browser trims the address and requires a complete value with content before and after @ plus a domain suffix. The full address is then checked one configured category at a time.

Source coverage

Account context, not mailbox verification

Only a row whose status normalizes exactly to Registered becomes a possible-account row. Many services expose registration signals through account-recovery behavior without revealing an account username or ID. A useful signal may therefore have no direct profile URL.

This workflow does not access the inbox, send an email, test delivery through SMTP, or prove that the address is active or controlled by a particular person.

05. API dispatch, pagination, and progress

1. Every scan request includes the identifier, one category, a page number beginning at 1, and only_found=false so non-positive rows remain available for review.
2. When multiple API base URLs are configured, the browser makes a fresh weighted selection for each request. Different categories or later pages can use different endpoints that expose the same response format.
3. The browser reads pagination.has_next after every response. Pages within one category remain sequential even when username categories run concurrently.
4. Reported scanned_modules values are accumulated against the configured total. The displayed percentage is held below 100% while requests remain, then closes at 100% after all category workers finish and no next page remains.
5. Each category progress bar accumulates returned source rows against that category's cached site count, including rows with negative, unknown, or error states. Site errors are kept in the response log. Select a category progress item to open the log at that category in its default partially expanded state. A failed page request keeps the progress already received and is marked Request failed; a category that returns fewer rows than its expected total is marked Incomplete. The response log retains the source, error, and page totals.

The pause control prevents the next page or category request from starting; it does not cancel a request already in flight. Restarting invalidates the earlier run in the interface so late responses from it are ignored. The client does not automatically retry a failed page request.

06. Source failures and uncertainty

External services control their own routes, response behavior, rate limits, regional rules, and anti-automation systems. Profile Finder displays the returned classification but does not independently re-fetch every source or override its rule.

07. Evidence and verification

A technical detection error and an identity-attribution error are different. A source can return the wrong kind of page, or it can return a real page that belongs to someone else. Both require human review.

Do not use one match, one missing result, or one visual similarity as the sole basis for identifying a person or making a consequential decision.

08. Privacy Mode and search history

Standard mode

The application creates a first-party history task before scanning. Exact positive matches can be normalized and saved as they arrive, and the task is marked complete after all category workers finish. Negative, unknown, and error rows remain in the live log and are not saved as positive matches.

Privacy Mode

Privacy Mode is off by default. When enabled before a scan, the browser skips first-party task creation, result saving, and task completion. The configured scan API still receives the username or email address, category, page, and request metadata required to perform the checks.

Google Dorks search shortcuts

After you submit a query, Google Dorks offers search links based on that exact username or full email address. Each shortcut explains its scope through a question-mark help button. Opening a link sends the generated search to Google in a new tab, including when Privacy Mode is enabled. These are independent searches of Google's index; they do not affect the scan progress or establish account ownership. You can inspect and copy every generated query before opening it.

Third-party requests and deletion

If a valid result URL is rendered, the browser may request a site icon from Google's favicon service. Stored individual history tasks and the current application account can be permanently deleted through the user workspace. See the Privacy Policy for the full data-flow and retention explanation.

09. Re-testing and corrections

Public-source checks age as platforms change. A route, API response, recovery flow, status vocabulary, pagination format, rate limit, or regional behavior can change without a visible change to this website.

To report a false positive, false negative, broken source, outdated category, or incorrect methodology statement, email [email protected]. Include the query type, affected service, observed status, expected behavior, approximate time, relevant public URL, and only the minimum safe reproduction detail. Never send credentials or private mailbox content.