find-anyone

find-anyone is an OSINT methodology skill for turning a person's name into a sourced, corroborated profile of one specific human. Its central premise is that the hard part is not collecting data but proving the data belongs to your subject rather than a namesake — so it front-loads disambiguation, requiring the investigator to bind the name to a second selector (employer, city, photo, email/handle, or age band) before collecting anything, and to define a discriminating test applied to every candidate record.

The workflow moves cheapest-and-highest-yield first: structured professional records and bios, then published output (bylines, papers with ORCID, patents), then regulated sources, with data brokers explicitly deprioritized because they hand back plausible wrong answers. An extensive source catalogue organizes references by the claim each can support — identity, employment and credentials, published output, business involvement, location and assets, and social/behavioral — noting each source's reliability and failure modes (e.g., self-asserted professional profiles, transcription errors in genealogy indexes, LLC-shielded property ownership).

The skill is explicitly framed for legitimate use — counterparty and investor due diligence, fraud and asset investigation, journalism, skip tracing, missing persons, pre-employment integrity checks, and self-exposure audits — and ships strong guardrails: a required authorized-scope step, a companion ETHICS.md, and an explicit always-out-of-bounds list (home address, movements, children, medical/immigration/sexuality status, and anything a reasonable person would call surveillance). Even so, the core capability is aggregating public records to profile a named individual, which is inherently dual-use and can be misused for stalking or doxxing, so it warrants elevated scrutiny despite its defensive framing.