The core operators
Every dork is a combination of these primitives. Chain them with parentheses and boolean keywords (AND, OR, -) to narrow results to the signal you care about.
Restrict results to a single domain or subdomain.
site:example.com
Match files by extension. Great for surfacing exposed configs.
site:example.com filetype:env
Match a word in the page title — useful for admin pages.
intitle:"index of" "backup"
Match a word in the URL path or query.
inurl:/admin/login
Match text inside the page body.
intext:"BEGIN RSA PRIVATE KEY"
Pull Google's cached copy of a page.
cache:example.com
Alias for filetype: — accepts same extensions.
site:gov ext:xls
Exclude results matching the keyword.
site:example.com -www
Recipe book
Replace target.tld with your in-scope domain. Run each query, document what's exposed, then write a short remediation note for the owning team.
Exposed environment files
site:target.tld (ext:env OR ext:ini OR ext:cfg) intext:"DB_PASSWORD"
Surfaces .env / config files accidentally deployed to public web roots.
Open directory listings
intitle:"index of /" site:target.tld
Misconfigured Apache/Nginx servers leak entire directory trees this way.
Login portals & admin panels
site:target.tld (inurl:admin OR inurl:login OR inurl:portal)
Maps attack surface for credential-stuffing or default-creds checks.
Public S3 / cloud buckets
site:s3.amazonaws.com "target"
Finds buckets indexed by Google — frequently contain backups and PII.
Leaked API keys on code hosts
site:github.com "target.tld" "api_key"
Public repos sometimes commit production keys; rotate on discovery.
Hands-on tasks
- Pick a public bug bounty program with a wide scope (HackerOne, Bugcrowd). Confirm Google Dorking is in scope before you start.
- Use
site:+filetype:to enumerate exposed documents (pdf,xls,doc). Note any internal naming patterns you find. - Map the org's login surface with
inurl:+intitle:. Do not attempt logins. - Write a 1-page report: queries used, assets surfaced, severity, suggested fix.
Stay legal
- Only dork assets that are in scope of a written authorization or a public bounty program.
- Finding exposed data via Google is passive; using credentials you discover is not — stop at discovery.
- Report findings to the owning team via the program's disclosure channel before publishing anywhere.
- The Google Hacking Database (GHDB) is a great reference but cross-reference current laws (CFAA, UK CMA, etc.).
Ready to pivot from passive to active?
Spin up a Kali VM and chain Google Dorking with active recon.
