| name | osint-and-secret-hunting |
| description | OSINT and secret-hunting tactics -- Google/GitHub/paste dorking, favicon-hash Shodan pivots, source-code disclosure paths (.git/HEAD, .env, .DS_Store, backup files), Wayback URL mining, steganography in uploaded images, metadata leaks, and WHOIS-derived username conventions. Converted from master-pentest-prompt.md Phase 32. Use early in an engagement, alongside recon, and again whenever a new asset (domain, image upload, employee name) surfaces. |
OSINT, stego & secret hunting -- real engagement tactics
When to use
Alongside recon at the start of an engagement, and again any time a new
asset surfaces mid-engagement -- a newly discovered subdomain, an image
upload feature, an employee name from a bio page.
Dorking
Google/GitHub/Bing operators: site: filetype: inurl: intitle:
intext:, inurl:q= for reflected-param discovery, Pastebin/Gist search
for company keywords, npm/pip package search by company name (dependency
confusion candidates), leaked creds on S3 (bucket-leak style search
tools).
Source code / config disclosure
.git/HEAD returning 200 (then dump the whole repo), .DS_Store
parsing (a dedicated ds-store tool), .env, .swp/.swo editor
leftovers, *~ backup files, .bak/.old backups, stray IDE project
files -- all are routine, high-yield checks on every target.
Historical / passive intel
- Wayback Machine URLs with
?q=/?api=/?token= params still baked
into old crawled links -- see the autonomous-research-loops skill's
Loop E for the deeper passive-asset-intelligence sweep.
- Public paste/breach dumps: HIBP, dehashed subsets, IntelX.
favicon.ico hash -> Shodan search (http.favicon.hash) to pivot
from one known asset to other infrastructure sharing the same
favicon.
Steganography
Both a CTF technique and something that shows up in real engagements
when employees embed creds in uploaded images: exif, strings,
binwalk, zsteg, steghide, stegsolve against any uploaded or
publicly hosted image.
Metadata leaks
EXIF GPS coordinates, author/editor names embedded in office documents
or images -- feeds directly into social-engineering pretexting even
when it isn't itself a technical finding.
Standard discovery files
robots.txt, sitemap.xml, humans.txt, security.txt, ads.txt --
hidden directories referenced in these routinely reveal admin panels,
staging environments, or an old parallel app.
Certificate transparency
CT logs surface internal hostnames via the issuer/SAN fields even when
those hosts were never meant to be public.
WHOIS / org intel
Registration data reveals org naming conventions, which in turn predicts
targeted usernames (firstname.lastname, flastname) for credential
stuffing or password-spray planning.