Skip to content
osint.platform

About

An OSINT platform that shows its working

Most lookup tools hand you a verdict. An investigation needs the opposite: the observation, where it came from, and how much weight it can carry. Every finding here carries its source, its confidence, and a plain statement of how far it can be trusted.

Responsible use

Built for cybersecurity research, OSINT research, defensive security, threat intelligence, journalism, academic research, authorised investigations, and personal digital footprint research. It uses public and legitimately accessible information only.

It performs no authentication bypass, no CAPTCHA solving, no private database access, and no access to any non-public profile. It is not built for stalking, harassment, doxxing, or credential theft. Platforms that cannot be checked without circumventing an access control are excluded by design, and the results say so.

Where the platform stops

These limits are deliberate. Each one is a place where a tool could produce a confident-looking answer that public evidence does not support, or could only get the answer by doing something it should not.

  • No password, token, cookie or private key is ever retrieved or displayed
  • No mailbox, account or login is ever probed with a credential
  • No email is ever sent to the address being investigated
  • No CAPTCHA or anti-automation control is ever circumvented
  • No port scanning, service probing or subdomain brute forcing
  • No private or paid personal-data broker is ever queried
  • No IP address is ever presented as a physical location
  • No malware sample is ever downloaded or executed

How evidence is treated

These rules are enforced in the code and the database, not just described here.

01

Nothing is fabricated

No provider invents a person, company, address, relationship, breach or source. Where a source has no answer, the platform says so rather than filling the gap.

02

A failed lookup is not a negative result

If a lookup errors, the finding reads 'could not be determined', not 'not published'. If a platform refuses the request, the answer is 'undetermined', not 'no account'. Absence of evidence is never dressed up as evidence of absence.

03

Disagreement is preserved

Reputation sources keep their own verdicts. When one feed says clean and another says malicious, both are shown and the confidence drops. Blending them into a single score would destroy the most useful thing an analyst can learn.

04

Identities are never merged on a guess

Entities are combined only on an exact match of a strong identifier: the same address, domain, IP or certificate fingerprint. Never a name, a display name or a photograph. A handle existing on several platforms is explicitly not treated as proof that one person holds them.

Reading a finding

Every finding carries one of four labels. They are never upgraded automatically, and the interface will not present a lead as a fact.

Verified
Observed directly from an authoritative source.
Inferred
Derived from other evidence by a documented rule, not observed directly.
Possible
A candidate that matches but is not corroborated. Treat it as a lead.
Unverified
Reported by a source that was not independently checked.

Your own research is private

Investigations are visible only to the session that created them. Nothing you research is shown to other users, sold, or sent to any advertising or analytics system. Investigations are deleted automatically after their retention period, and you can delete any of them immediately.