Start with an exact email address. PrufAgent searches browser-visible public references, possible reused usernames, profile pages, websites, and identity details, then shows a masked preview before any one-time report purchase.
This is a real PrufAgent email preview capture. It separates checked coverage, possible public candidates, source evidence, exposure context, and limitations so an AI summary cannot masquerade as a verified social profile.
Switch modes above to see the username, name, or phone product view. A paid email report can include cited public URLs, evidence notes, source-quality labels, identity-confidence labels, conflicts, and checked-source limits.
An email can be distinctive, but it may be old, shared, forwarded, compromised, used for a business, or disconnected from the person you are researching. Each candidate needs context from the source page.
The complete address may appear on a public profile, contact page, business site, portfolio, forum, public document, code repository, or indexed post. Exact text is useful evidence that the page mentioned the address, but it does not establish current control.
The text before the @ sign can suggest a username to check on public pages. Common names, initials, and generated strings create false candidates, so a derived handle stays weak until independent details agree.
A personal website, organization page, creator profile, or public social page may connect the address to a name, location, biography, domain, organization, or another identifier.
A candidate may have the wrong location, unrelated work history, different linked domain, or incompatible biography. A useful report preserves those conflicts instead of forcing every hit into one identity.
The first pass uses the complete normalized address because exact references are less ambiguous than a common display name. The engine can look for public pages that repeat the address and identify surrounding names, aliases, organizations, websites, and profile links. The domain also supplies context: a personal provider, school, employer, organization, custom domain, and disposable provider each imply a different research path.
The second pass treats the local part as a candidate handle, not a confirmed account. A local part such as first.last may map to many unrelated people. A distinctive alias that appears beside the same website, location, biography, or profile image can be more useful. A username available in a social URL is not evidence that the email owner controls it.
The final pass compares independent details across sources. A consistent name, location, organization, linked website, public biography, image, and second verified identifier can raise confidence. An old date, copied directory entry, unrelated biography, or conflicting location must lower it. Source agreement matters more than the number of search-result snippets.
AI broadens discovery, compares candidate identities, and organizes evidence. It does not bypass authentication, and model output alone is never accepted as proof.
The application validates and normalizes the submitted email, keeps the exact clue within the report flow, and generates separate public-search candidates only where the evidence path supports them.
Available browser-visible pages and structured providers are checked. A source page, search snippet, redirect, generic landing page, block, deletion, and login wall remain different outcomes.
OpenRouter-routed models search for relevant references, compare candidate identities, identify conflicts, and structure source-backed findings for the report.
The report presents cited URLs, evidence notes, source quality, confidence, conflicts, and limitations. Unconfirmed candidates remain unconfirmed instead of becoming facts through repetition.
The underlying public source, not model output alone, is what can support a finding. A profile URL, public page, document, or indexed public reference should carry a clear evidence note explaining what it does and does not show. A model can propose a search query or candidate. It cannot turn that candidate into verified ownership without corroboration.
PrufAgent also distinguishes source quality. A first-party public profile or organization page generally carries more weight than a copied directory, scraped list, old search snippet, or anonymous post. Dates matter because people change employers, domains, usernames, locations, and accounts.
When providers fail or disagree, the report should show the limitation. A missing result is preferable to invented certainty. That evidence discipline makes the AI useful for real research rather than merely confident-sounding text.
Social platforms control what profiles, fields, and pages are public or indexed. PrufAgent can organize browser-visible evidence; it cannot promise a private email-to-account mapping that a platform does not expose.
| Evidence type | What it can support | What it cannot support alone | PrufAgent treatment |
|---|---|---|---|
| Exact email on a public first-party profile | Strong page-level association | Current control, intent, or complete identity | Cite the page, capture context, date it, and compare another identifier. |
| Email on a personal or business website | Connection to that public site or organization | Ownership of every linked social account | Follow public profile links and compare names, domains, bios, and locations. |
| Local-part username candidate | A possible handle worth checking | That the email owner controls a matching username | Label as candidate-only until independent public details corroborate it. |
| Copied directory or search snippet | A lead to locate a better source | Freshness, accuracy, or current profile ownership | Seek the underlying page and preserve staleness or copy risk. |
| Private, blocked, or login-only page | Only that public access was unavailable | Profile contents, membership, or account ownership | Record the limitation. Do not infer hidden details or bypass access controls. |
Facebook documents that people can control who sees an email address on a profile. It also explains that usernames and user IDs are part of the public profile and can help people find a profile. Those are separate facts: a public username does not prove that a submitted email controls the account.
Review Facebook's guidance on email visibility and public usernames and user IDs.
LinkedIn documents settings for who can see a member's email address. A public LinkedIn profile may expose a name, headline, organization, location, public activity, or linked site without exposing the submitted email. Search results therefore need contextual correlation, not a claim that LinkedIn disclosed a private registration address.
Review LinkedIn's current email visibility guidance.
The same public-evidence rule applies across platforms. A public bio, profile URL, linked website, creator contact page, public post, or organization page can be reviewed. Private profiles, app-only content, personalized recommendations, direct messages, hidden registration data, and login-walled pages are outside the public report's promise.
Platforms also change indexing, page structure, access controls, and privacy defaults. A method that worked previously may stop working, and a public profile may be renamed or deleted. PrufAgent should report what was checked now, provide source URLs when available, and avoid implying permanent coverage.
Search engines can display copied or stale pages long after a profile changes. Open the cited page, compare its current content, and check dates. The absence of a profile from a search engine does not prove absence from the platform.
A useful social profile search is a sequence of checks. Each step should either add corroboration, reveal a conflict, or establish a clear limit.
Begin with the complete address and inspect the page around every exact occurrence. Record the page title, source URL, date, surrounding name or organization, and whether the source is first-party, copied, or user-generated.
Extract a possible handle only when the local part is meaningful. Check dots, underscores, initials, and obvious separators conservatively. Do not generate a large list of speculative identities from a random string.
Review public biographies, profile images, locations, work or school details, linked websites, dates, and recurring aliases. Require agreement across independent details before raising confidence.
Do not discard a mismatched location, occupation, timeline, organization, or image. A contradiction can show that two matching usernames belong to different people or that a directory entry is stale.
First-party public pages deserve the closest review: a public social profile, personal website, employer page, organization directory, or creator contact page. A page controlled by the subject or a named organization can provide direct context, though it may still be outdated or impersonated.
Secondary reporting, conference pages, public documents, archived pages, and reputable directories can support context. Copied aggregators, scraped profile lists, unattributed databases, and search snippets should be treated as leads because they can reproduce one bad source many times.
Count independent evidence, not duplicated text. Ten sites copying the same directory record are still one underlying claim. A good report should identify repeated content and prioritize the original source where possible.
This page starts with email because that is the query you brought. The same tool can route a username, name with place, or phone through its correct preview and report path.
Best for exact public address references, possible aliases, linked websites, public profiles, and source-specific exposure context. Old references do not prove current control.
Report: Public Signal, $9.99 once
Best for cross-platform public handle trails. Identical usernames require biography, location, image, website, organization, or another verified identifier before confidence rises.
Report: Public Signal, $9.99 once
Best for contextual discovery when you have a verified city, region, employer, school, organization, profession, or business. Common names need more context.
Report: Public Signal, $9.99 once
Best for supported number structure and exact-number public references. Numbers can be reassigned, shared, ported, or published for a business.
Report: Phone Clue, $4.99 once
If the email search, likely username search, and name-with-location search point to the same website, organization, biography, and public profile, the combined trail may be more coherent. If the phone points to another region or the username has an unrelated biography, preserve the conflict.
Do not enter guessed emails simply to produce more results. A fabricated address creates noise and can implicate an unrelated person. Use clues supplied directly, published publicly, or otherwise obtained for a legitimate purpose.
When two candidates remain plausible, keep both and label the uncertainty. A report that shows two unresolved people is more useful than a polished but unsupported choice.
The masked preview runs first. The app then shows the selected report type, exact price, included evidence, and one-time purchase disclosure before opening Stripe.
For email, username, or name-with-context searches. The report can include public candidates, cited URLs, evidence notes, quality labels, confidence, conflicts, and limitations.
For an exact phone clue. The report checks supported number structure and public exact-number references, then labels only what checked sources support.
An exact email can surface public web references, possible reused usernames, public profiles, websites, aliases, and identity clues when sources expose them. It cannot guarantee access to a private, blocked, deleted, or login-only profile.
PrufAgent searches browser-visible public sources and cites supported profile candidates. It does not claim private platform APIs, member-database access, account recovery access, or a way around platform privacy settings.
It can reveal exact public email mentions, possible handle reuse, public profile URLs, personal or business websites, aliases, biographies, locations, organizations, and conflicts between candidates.
No. A username derived from the part before the at sign is a discovery clue. Confidence improves only when independent details such as name, location, biography, image, linked website, organization, or another verified identifier agree.
It means the checked public sources did not support a stronger finding at that time. It does not prove that no private, renamed, deleted, blocked, or unindexed social profile exists.
The masked preview is free. The Public Signal Report costs $9.99 once and is not a subscription. The exact product and price are shown before Stripe checkout.
No. PrufAgent is not a consumer reporting agency and must not be used for employment, tenancy, credit, insurance, or another regulated eligibility decision.
Focus an email search on browser-visible Facebook profile clues and Facebook's privacy limits.
Find Facebook clues by emailSearch broader public identity, website, organization, profile, and exposure context around an address.
Open reverse email lookupUse an exact phone clue when the email is unavailable or produces sparse public evidence.
Search social profiles by phoneFollow a likely handle across browser-visible public profiles while preserving identity conflicts.
Find someone by usernameFollow a clue-by-clue workflow for public profiles, source review, and identity confidence.
Read the social search guideUse the dating-focused page when relationship context and dating-adjacent public references matter.
Search dating clues by emailEnter the email, review the masked public signal shape, and see the exact $9.99 one-time report price before checkout.
Search an email
Use only for lawful public research. Do not use PrufAgent for stalking, harassment, or regulated eligibility decisions.
Free masked preview. The selected one-time price appears before checkout. No subscription.