google-like-a-spy

This OSINT skill teaches advanced search-engine query craft and Google dorking to surface hidden files, documents, and mentions. It explains how operators act as filters over a crawled index, why coverage and operator fidelity differ between Google, Bing, DuckDuckGo, and Yandex, and how to force literal matching so a query runs as typed rather than being silently stemmed or synonym-expanded. It documents which operators still behave (exact phrase, site:, filetype:/ext:, intitle:/inurl:/intext:, before:/after:, exclusion and alternation) and which retired ones no longer earn their place.

The reference material is organized by objective. A dork cookbook provides ready-to-adapt query patterns for finding an organization's documents (including copies hosted on third-party and government sites), enumerating a person's accounts and mentions by selector rather than name, and locating a leaked selector across paste and code hosts. It also includes an explicitly dual-use section for surfacing exposed files: directory listings, configuration and environment files, database and backup artefacts, cloud object-storage buckets, credential-shaped content such as private-key and API-key markers, and admin interfaces that should not be public. The skill wraps this with strong ethical framing, repeatedly stating that discovery is only for estates you own or are contracted to assess, that finding a misconfiguration is passive but fetching or using it is not, and directing findings to a reporting route.

Target users are OSINT practitioners, security researchers, and investigators doing data-exposure audits, pre-engagement reconnaissance, competitive or regulatory research, and insider-leak investigation. Because the same techniques that support authorized audits also enable locating other parties' exposed credentials, backups, and buckets, this is inherently dual-use reconnaissance tooling despite its ethics guidance.