0% Complete
0/0 Steps
  1. About

    About Cyber School
  2. About BDO
  3. The CSRP program Syllabus
  4. Module 1
    The Cyber World and CIA Triad
    6 Topics
  5. The OSI Model and Network Layers
    6 Topics
  6. IP Addressing, Subnetting and Routing
    6 Topics
  7. DNS and DHCP
    6 Topics
  8. Windows Server Administration
    6 Topics
  9. Active Directory and Identity Management
    6 Topics
  10. Linux CLI Fundamentals
    6 Topics
  11. Linux Permissions and Hardening
    6 Topics
  12. Module 2 - Defense Technologies and Cyber Tools
    Firewall Architecture and Rule Design
    5 Topics
  13. NAT and Secure Traffic Mapping
    6 Topics
  14. Enterprise VPN Solutions
    6 Topics
  15. IDS and IPS
    5 Topics
  16. Identity and Access Management
    5 Topics
  17. Multi-Factor Authentication
    6 Topics
  18. Data Loss Prevention
    6 Topics
  19. Endpoint Security and EDR
    6 Topics
  20. Mail Security
    6 Topics
  21. Web Proxy and URL Filtering
    6 Topics
  22. Regulation and Compliance
    6 Topics
  23. Cloud Security
    6 Topics
    |
    1 Quiz
  24. Module 3 - Ai in security
    AI in Security
    5 Topics
  25. ML Anomaly Detection
    5 Topics
  26. AI Threats
    5 Topics
  27. LLM Defense
    6 Topics
  28. Module 4 - Offensive Cyber & Vulnerability Scanning
    Cyber Kill Chain and Attacker Mindset
    6 Topics
  29. OSINT and Public Intelligence Gathering
    6 Topics
  30. Network and Port Scanning with Nmap
    6 Topics
  31. Vulnerability Scanning
    6 Topics
  32. Exploitation Fundamentals
    6 Topics
  33. Web Application Security
    5 Topics
  34. Social Engineering
    5 Topics
  35. Password and Hash Cracking
    5 Topics
  36. Module 5 - SOC and Advance Cyber Analyst
    SOC Structure and Operations
    6 Topics
  37. SIEM Fundamentals with Splunk
    6 Topics
  38. Alerting and Dashboards
    6 Topics
  39. Incident Response
    5 Topics
  40. Basic Digital Forensics
  41. Capstone A - The Attack: Breaching Meridian
  42. Capstone B - The Defense: Defending Meridian
  43. ISC2 CC Mock Exam and CSRP Graduation
    2 Quizzes
Lesson Progress
0% Complete

CSRP  —  Cyber Security Responder & Practitioner

Lesson 26: OSINT and Public Intelligence Gathering

Textbook

Cyber School  ·  BDO Ziv Haft

CSRP LESSON 26 — TEXTBOOK

Module 4 — Offensive Cyber & Vulnerability Analysis · Lesson 2 of 8 · Course Lesson 26 of 40 Title: OSINT and Public Intelligence Gathering Deep-reading companion to the Gamma deck. Read before or after the live session.


⚠️ Ethical-use contract

Everything in Module 4 is taught for defensive purposes. The techniques in this textbook are real techniques real attackers use. You will only practice them in authorized lab environments. You will not run dig, whois, exiftool, or any active recon tool against any real organization without written authorization.

Outside this contract: – Passive recon is generally legal, but specific techniques (Whois at scale, LinkedIn scraping, certificate transparency at scale) can violate terms of service and rate limits. – Active recon (port scans, vulnerability scans, banner grabbing) can constitute unauthorized access under the US Computer Fraud and Abuse Act, the UK Computer Misuse Act, Israel’s Computer Law, and equivalent legislation in every country we operate in. – Real-world penetration testing requires written authorization, defined scope, and rules of engagement.

When in doubt — ask before you act.


TABLE OF CONTENTS

  1. The concept of Open Source Intelligence
  2. What counts (and what doesn’t) as OSINT
  3. Passive vs active reconnaissance — the boundary that matters
  4. The legal landscape — CFAA, Computer Misuse Act, Israel’s Computer Law
  5. OSINT sources — a practitioner’s inventory
  6. DNS as OSINT — records, posture, exposure
  7. Whois — the registration-metadata layer
  8. Certificate transparency — the subdomain reveal
  9. Document metadata — the forgotten leak
  10. Search engines, GitHub, archives — the high-leverage sources
  11. OSINT for defenders — turning the lens on yourself
  12. The defender’s quarterly OSINT cadence
  13. How OSINT feeds the rest of Module 4
  14. Cross-references and further reading

1 · The concept of Open Source Intelligence

Open Source Intelligence — OSINT — is the discipline of gathering information about a target using publicly available sources. “Open source” here is the intelligence-community sense, not the software sense: it means unclassified, public-facing information, not “open-source software.” The term predates computer security. Intelligence agencies used OSINT throughout the 20th century to harvest information from newspapers, broadcast radio, public registries, academic publications, and trade journals.

In the cyber security context, OSINT is the first phase of nearly every targeted attack. It is also the first phase of nearly every disciplined defender’s quarterly exposure review. Reconnaissance is Phase 1 of the Lockheed Martin Cyber Kill Chain you mapped in L25, and OSINT is the operational practice within that phase — the techniques and tools that produce the target dossier that drives every later phase.

> KEY TERM — OSINT: the structured gathering of intelligence about a target from publicly available, unclassified sources. NOT covert. NOT exploitative. NOT requiring authorization to consume (though the act of consuming at scale may violate terms-of-service).

The reason OSINT matters so much in modern attacks is that the cost-to-benefit ratio is overwhelmingly favorable to the attacker. Reading a target’s public website costs nothing, leaves no traffic on the target’s logs, produces no IDS alerts, and generates an immediate dossier of names, roles, technologies, vendors, projects, and exposed services. The attacker has spent two hours and has a dossier the defender’s CISO would have struggled to assemble. That asymmetry is what L26 teaches you to invert — by running the same techniques on your own organization, before the attacker does.

> REAL-WORLD — An IBM Security X-Force Threat Intelligence Index analysis of major incidents in the early 2020s found that the majority of targeted intrusions began with extensive OSINT-driven reconnaissance: harvested LinkedIn profiles, forgotten subdomains from CT logs, leaked PDFs with author and internal-path metadata, exposed credentials in public GitHub commits. The technical sophistication of the initial-access step is often low; the intelligence quality is what makes the attack succeed.


2 · What counts (and what doesn’t) as OSINT

The most common confusion among new practitioners is the line between OSINT and active reconnaissance. The line is observability from the target’s perspective.

What counts as OSINT (passive, no traffic to target)

  • Company websites + employee bio pages + press releases + careers / job postings
  • LinkedIn profiles + Twitter/X / Mastodon / Reddit profiles + employee social media
  • DNS records (queried via your own resolver or a third-party recursive DNS service — the target’s authoritative server may eventually see the query indirectly, but the attacker’s identity is not associated with it)
  • Whois records (queried via third-party services that broker the registry response)
  • Certificate transparency logs (public, append-only, audited by CT log operators)
  • Public document metadata (PDFs, DOCX, images that the target has chosen to publish)
  • Search engine results (Google, Bing — including Google dorking with operators like site:, filetype:, inurl:)
  • Public code repositories (GitHub, GitLab) — committed secrets, configs, internal hostnames
  • Public archives (Internet Archive Wayback Machine, certificate transparency aggregators like crt.sh)
  • Public threat intelligence feeds (some commercial, some open — VirusTotal, AbuseIPDB, Shodan, Censys)

What does NOT count as OSINT

  • Active probing of systems (port scans, vulnerability scans, banner grabbing — L27 and L28 territory)
  • Login attempts even with default or guessed credentials (L31, L32)
  • Anything requiring authentication or breach of a system
  • Anything covered by an NDA or platform terms-of-service violation
  • Probing any system without written authorization

> TIP — If a journalist could obtain the information without breaking the law, it’s OSINT. If a journalist would need a court order, a leaker, or a hacked source, it’s not.

The “grey zone”

Some sources sit at the edge of passive:

  • Certificate transparency logs are public, but querying them at scale is observable to the CT log operators (e.g., crt.sh runs analytics on its query patterns).
  • Search engines log queries; Google Search Console can show the queries that lead to a domain. So while the target doesn’t see the attacker querying, the target may indirectly see the queries.
  • The Wayback Machine is passive when you read snapshots, but creating a new snapshot for a target’s site is closer to active — your archiver fetches the target’s pages.
  • whois via the registrar’s WHOIS service is generally passive, but at scale it triggers rate-limits and can lead to IP blocks.

The CSRP convention: when in doubt, stay clearly passive AND limit yourself to authorized targets. The habit matters as much as the technicality.

> WATCH OUT — In some jurisdictions (and under some platform terms), even scraping a public LinkedIn profile programmatically has been litigated and lost. hiQ Labs v. LinkedIn (US) eventually established broad protection for scraping public profile data under CFAA, but the platform’s terms-of-service can still be enforced via contract law. Read the terms before you script anything.


3 · Passive vs active reconnaissance — the boundary that matters

The distinction is best understood through the target’s logs.

Passive recon

The attacker reads the target’s public surface via intermediaries: – The attacker’s browser fetches https://target.example/about.html — the target’s web server sees a single hit, indistinguishable from any other visitor. There is nothing to attribute back to “an attacker.” – The attacker queries Whois via whois.iana.org or a third-party Whois service — the target’s domain registrar logs a Whois query but does not know who the requester is. – The attacker queries the public CT log crt.sh for target.example certificates — the CT log operator (Cloudflare, Sectigo, Google) logs the query; the target itself never sees it.

The defender’s perspective: passive recon is invisible. You cannot detect it on your own infrastructure. You can only infer that it’s happening by the fact that an attack inevitably follows in some unknown future. Which means the defensive mitigation for OSINT is NOT detection — it is exposure reduction. You shrink what the attacker can read.

Active recon

The attacker generates packets the target can observe: – A port scan against target.example lights up the target’s firewall logs with a series of SYN packets. – A DNS zone transfer attempt (AXFR) hits the target’s authoritative DNS server. – A banner grab on TCP/22 fetches the SSH banner — the target’s SSH daemon logs the connection.

The defender’s perspective: active recon is detectable. Firewall logs, IDS signatures, reputation feeds, and anomaly detection can all surface active scanning patterns. The defensive mitigation for active recon includes both detection (alert on it) and prevention (rate-limit, block, geofence).

> KEY TERM — Passive recon: information gathering that does not generate observable traffic at the target. Active recon: information gathering that does generate observable traffic at the target.

Why attackers stay passive

A sophisticated attacker stays passive as long as possible: 1. Operational security: every active probe is a chance to be caught. Passive recon doesn’t leave fingerprints. 2. Defender preparation time: the longer the attacker stays invisible, the less time the defender has to harden in response. 3. The dossier is often enough: for credential-stuffing campaigns, phishing, or social engineering, a thorough OSINT dossier is sometimes the only reconnaissance the attacker needs.

Why attackers eventually go active

Some attacks require active recon: knowing which ports are open, which services are running, which versions of which software respond. The attacker will eventually trigger detectable traffic — but a disciplined operator minimizes the active footprint, using slow scans, decoys, and credentialed access where possible.

L26 is passive only. L27 introduces active scanning with Nmap. The pedagogical sequence — passive then active — mirrors how a real attacker (and a real penetration tester) sequences engagement.


4 · The legal landscape

The legal status of OSINT and reconnaissance varies by jurisdiction. The CSRP-relevant frameworks:

United States — Computer Fraud and Abuse Act (CFAA, 18 U.S.C. § 1030)

The CFAA prohibits “intentionally accessing a computer without authorization or exceeding authorized access.” The Supreme Court’s 2021 decision in Van Buren v. United States narrowed the “exceeding authorized access” prong substantially — making purely-public-data scraping much harder to prosecute under CFAA. But the unauthorized-access prong remains broad. Active scanning against a non-authorized target is risky. Passive consumption of intentionally-public information generally is not.

United Kingdom — Computer Misuse Act 1990

The Computer Misuse Act criminalizes “unauthorized access” to a computer (Section 1) and broader unauthorized acts. UK case law (e.g., DPP v Bignell) has held that even exceeding authorized access can constitute an offense. UK practitioners are advised to be conservative about active recon without written authorization.

Israel — The Computer Law (חוק המחשבים, 5755-1995)

Israel’s Computer Law criminalizes unauthorized intrusion into computer materials (Section 4 of the Computer Law). The standard for “intrusion” is set by the Israeli Supreme Court at a relatively low threshold compared to some other jurisdictions. Active scanning against an Israeli target without authorization is a real legal risk; passive consumption of public information is not.

European Union — NIS2 Directive + national laws

The EU does not have a single hacking statute. Each member state’s national law applies (e.g., Germany’s StGB §202a-c, France’s Code pénal articles 323-1 to 323-3). The pattern is similar: unauthorized access is criminalized; active probing without authorization is risky.

> KEY TERM — Written authorization: a documented permission from a target organization’s appropriate authority (e.g., the CISO, General Counsel) granting a specified individual or team the right to perform specified security-testing activities against specified systems within a specified time window. The document defines scope, target list, allowed techniques, and rules of engagement. Real penetration testing engagements always have one. CSRP labs are an internal authorization granted by the BDO Ziv Haft engagement contract.

The CSRP rule

In every CSRP lesson and in every CSRP graduate’s professional practice: – Lab environments only for offensive techniques. – Written authorization for any real-target engagement after graduation. – Defensive intent for every offensive skill acquired.


5 · OSINT sources — a practitioner’s inventory

A working OSINT practitioner draws from several source families. The list below is approximately ordered by “first-look” priority for most engagements.

5.1 · The target’s own public website

The richest single source. Staff bios reveal names, roles, and email patterns. Careers / job postings reveal technology stack with extraordinary specificity (because they’re written to attract qualified candidates). Press releases reveal partnerships, acquisitions, customer wins. White papers and datasheets reveal product capabilities. Privacy policies and terms of service reveal which subprocessors and third parties handle data.

> TRY IT YOURSELF — Open your own organization’s public website tomorrow morning. Browse it as if you were an attacker. List every employee name, every email shown, every technology mentioned. The exercise will surprise you.

5.2 · LinkedIn

Employee inventory at scale. Job titles, tenures, claimed responsibilities, endorsements (which reveal claimed skills and tools). Some practitioners argue LinkedIn is the single highest-value OSINT source for target enumeration in 2025-2026 attacks.

5.3 · DNS

A / AAAA / MX / TXT / CNAME / NS / SOA records. Cloud-provider IP-range identification. SPF/DMARC/DKIM posture. Subdomain enumeration via certificate transparency. We dedicate a chapter to DNS (Chapter 6 below) because the toolkit is rich and the defender’s leverage is high.

5.4 · Certificate Transparency logs

Every certificate issued by a publicly-trusted CA since 2018 has been logged to a public, append-only CT log. Search tools like crt.sh, censys.io, and certificate.transparency.dev reveal every subdomain anyone has ever requested a certificate for — including internal-staging subdomains that were briefly internet-facing.

5.5 · Public document metadata

The chapter-4 focus. PDFs, DOCX, JPGs, sometimes PNGs and SVGs. Author names, internal file paths, software versions, GPS coordinates, comments, tracked changes. Marketing publishes a PDF; the metadata reveals the internal user who wrote it, the internal file-share path, and the software version (which may have a CVE).

5.6 · Search engines + Google dorking

Operators that narrow search results: site:target.example filetype:pdf, inurl:admin, intitle:”index of”, cache:. The Google Hacking Database (GHDB at exploit-db.com) catalogs thousands of dorks for finding misconfigured services, exposed credentials, and forgotten admin pages.

5.7 · Public code repositories

GitHub, GitLab, Bitbucket. Searches for the target’s domain name, internal hostnames, and common secret patterns (API keys, AWS access keys, private keys, database connection strings). Free tools: truffleHog, gitleaks, GitHub native secret scanning.

5.8 · Public archives

The Internet Archive’s Wayback Machine has petabytes of historical web content. Sometimes content the target has deleted is still archived. Combined with current-state OSINT, the archive reveals how the target’s public posture has changed — including roles, technologies, and services that were exposed in the past but no longer are (or vice versa).

5.9 · Threat intelligence feeds + passive infrastructure databases

  • Shodan — internet-wide port-scan data; reveals public-facing services indexed by version, location, and protocol. Often catches misconfigured or forgotten cloud resources.
  • Censys — similar to Shodan with stronger certificate-data integration.
  • VirusTotal — passive DNS, file-hash analysis, related-domain enumeration.
  • AbuseIPDB / AlienVault OTX — known-bad IP reputation and threat intel.

5.10 · Have I Been Pwned

The HIBP service indexes credential breaches. For an organization, querying its email domain reveals which employees have been in past breaches — informing both phishing pretexting (for the attacker) and credential-reset campaigns (for the defender).


6 · DNS as OSINT

DNS is public by design. It is also the second most useful single source (after the target’s website) for the typical engagement. The toolkit:

ToolUse
digMost flexible. Query specific record types with options for trace, short output, alternate DNS servers.
nslookupOlder, simpler. Interactive mode for quick queries.
hostSimpler than dig for one-shot queries.
whoisDomain registration metadata.
crt.sh (web)Certificate-transparency log search.
dnsenum / dnsrecon / amass / Sublist3rSubdomain enumeration via multiple passive sources.

Records and what they reveal

  • A / AAAA — the IPv4 / IPv6 address(es) of the host. Reveals cloud provider (AWS/Azure/GCP IP ranges), CDN (Cloudflare, Akamai, Fastly), or on-prem. ASN lookup (whois -h whois.cymru.com “AS<ip>”) maps the IP to the routing entity.
  • MX — mail-server entries. Reveals mail vendor (Google Workspace = aspmx.l.google.com; Microsoft 365 = *.mail.protection.outlook.com; Proofpoint = mxa-*.mxrecord.com).
  • TXT — text records. Most commonly used for SPF (v=spf1 …), DMARC (v=DMARC1 …), DKIM, and domain-verification tokens for various SaaS services.
  • CNAME — aliases. Subdomain → external host mappings. A dangling CNAME (pointing to a cloud resource that has been deleted) is a subdomain takeover vulnerability.
  • NS — authoritative DNS servers. Reveals which DNS provider hosts the zone.
  • SOA — Start-of-Authority. Includes the zone admin’s email (technically — often a placeholder).

SPF/DMARC parsing

SPF (Sender Policy Framework, TXT record starting v=spf1) declares which mail servers are authorized to send email on behalf of the domain. The terminating mechanism is the policy: – +all — permissive; ANY server can send. A misconfiguration. – -all — strict; only listed servers can send. Failing email is rejected. – ~all — soft fail; failing email is marked but typically delivered. – ?all — neutral; no policy. Defensively weak.

DMARC (TXT record at _dmarc.target.example, starting v=DMARC1) tells receiving mail servers what to do with email that fails SPF or DKIM: – p=none — monitor only; do nothing to failing email. Defensively weak. – p=quarantine — send failing email to the spam folder. – p=reject — bounce failing email entirely. Strongest posture.

> WATCH OUT — A publicly-discoverable p=none DMARC is a green light for spoofers. Attackers query _dmarc.target.example BEFORE crafting phishing campaigns, specifically to confirm the target’s policy.

Certificate transparency — the underrated reveal

Modern CAs (Let’s Encrypt, DigiCert, Sectigo) log every certificate they issue to public CT logs. Querying crt.sh for a target domain reveals every subdomain anyone has ever requested a certificate for — including subdomains the target has long forgotten:

  • dev-staging.target.example from a 2019 cloud experiment that’s now an abandoned Azure resource.
  • internal-admin.target.example from a 2021 misconfiguration that briefly exposed an admin panel.
  • acme-challenge-*.target.example artifacts from automated cert issuance.

Defender implication: assume every subdomain you have ever requested a cert for is known to attackers. Inventory + harden accordingly.


7 · Whois — the registration-metadata layer

Whois is the protocol for querying domain registration metadata from the registry / registrar. Returns: – The registrar of record – Registration date + expiry date – Name servers – (Sometimes) the registrant’s contact email and organization name — increasingly hidden by privacy services

> TIP — Whois data is most useful for organizational pivoting. Two domains registered on the same day with the same name servers from the same registrar often belong to the same organization. Whois history services (DomainTools, RiskIQ) preserve old Whois records even after the registrant adds privacy protection.

The GDPR era (2018+) substantially reduced the personal-data fields in Whois. Most modern Whois responses are anonymized for EU-registrant domains. For non-EU domains, more data may remain visible.


8 · Certificate transparency — deeper

CT logs are append-only, public, cryptographically verifiable logs of every TLS certificate issued by a publicly-trusted CA. The CT ecosystem was established to defend against rogue or compromised CAs issuing fraudulent certificates. The side-effect — every subdomain anyone has ever requested a cert for is publicly logged — is what makes CT a top-tier OSINT source.

Querying CT logs: – crt.sh — the most popular public interface. Supports wildcard searches: %.target.example. – censys.io — programmatic access via API; richer certificate-data integration. – certificate.transparency.dev — the Google/Sectigo CT log search portal.

For defenders: querying crt.sh for your own domain quarterly is the single best way to discover forgotten subdomains. Many incidents have started with attackers finding a forgotten subdomain in CT, hitting it, and finding it pointed at a long-abandoned cloud resource that the attacker can now take over.


9 · Document metadata — the forgotten leak

Office documents (DOCX, PPTX, XLSX), PDFs, and images all carry metadata. The metadata may include:

PDFs

FieldWhat it reveals
AuthorThe user who authored the file
CreatorThe application that generated the source document
ProducerThe PDF-generation tool + version
CreationDate / ModifyDateTimestamps
Title / Subject / KeywordsSometimes left from a template
CommentsAuthor-added comments (sometimes sensitive)
EmbeddedFilesOther files embedded inside the PDF

DOCX / PPTX / XLSX

Office files are ZIP archives containing XML. The metadata lives in docProps/core.xml and docProps/app.xml:

FieldWhat it reveals
Author / LastModifiedByUser identities
ApplicationOffice version
CompanyOrg name
Lines / Words / PagesContent statistics
Revision (and track-changes history)Editing history — sometimes leaks earlier drafts
CommentsAuthor-added comments

Images (JPG, PNG)

FieldWhat it reveals
Make / ModelCamera or phone model
SoftwarePhoto-editing tool + version
DateTimeWhen the photo was taken
GPS Latitude / Longitude / AltitudeWhere the photo was taken (phones with GPS enabled)
Camera serial numberSometimes — links photos across leaks

exiftool — the universal metadata reader

exiftool reads metadata from ~100 file formats. Key commands:

exiftool <file>                           # dump all metadata
exiftool -a -G1 <file>                    # group by metadata family
exiftool -CommonTags <file>               # only high-signal fields
exiftool -GPS* <file>                     # only GPS fields
exiftool -Author -Creator <file>          # specific fields
exiftool -all= -overwrite_original <file> # strip ALL metadata (irreversible)

> REAL-WORLD — In 2013, a US Department of Defense declassified PDF revealed, in its metadata, the original DoD employee author and the internal file-share path. Investigators who knew to check exiftool extracted those details from a document that had been “officially” redacted. Marketing-published PDFs leak the same way — only with smaller national-security consequences and larger competitive ones.

Defender’s response — the metadata-stripping pipeline

For PDFs: qpdf –linearize –object-streams=generate <input> <output> strips most metadata. Or exiftool -all= -overwrite_original. For Office-generated PDFs, use the “Inspect Document” → “Remove” before saving.

For DOCX/PPTX/XLSX: Office’s built-in “Document Inspector” (File → Info → Inspect Document → Remove All). Or, in a CI pipeline, convert documents to PDF via a metadata-stripping pipeline before publication.

For JPGs: exiftool -all= -overwrite_original strips EXIF including GPS. For phone-camera photos especially.

The mature posture: a pre-publication metadata-stripping pipeline integrated into the publishing workflow, so that any document leaving the organization has been scrubbed automatically.


10 · Search engines, GitHub, archives — the high-leverage sources

Google dorking

Specific operators that narrow search results:

OperatorPurpose
site:target.exampleRestrict to a domain
filetype:pdfRestrict to PDF results
inurl:adminURL contains “admin”
intitle:”index of”Page title contains “index of” (directory listings)
cache:target.example/pageGoogle’s cached copy
intext:”password”Page body contains “password”

Combined, dorks find misconfigured services and exposed documents: site:target.example filetype:xlsx may surface every published spreadsheet on the target’s domain.

GitHub exposure

GitHub native secret scanning automatically detects committed AWS keys, GitHub PATs, and other common secret patterns — but only for repos in organizations that have enabled it. Public commits across the global GitHub corpus are searchable via the GitHub Search API.

Tools: – truffleHog — historical-commit secret scanner with entropy-based detection. – gitleaks — similar, with regex-based detection. – GitHub native secret scanning — built-in for enabled repos.

Defender’s quarterly task: search public GitHub for your organization’s domain name, internal hostnames, and common secret patterns. Free, fast, high-signal.

The Wayback Machine

web.archive.org archives web content. Querying snapshots of target.example from 2015, 2018, 2021 may reveal historical content that no longer exists on the live site:

  • Old employee directories with people who have since left.
  • Old technology mentions (“powered by [outdated tool with known CVE]”).
  • Old subdomain listings.
  • Old privacy policies (sometimes useful for legal discovery).

11 · OSINT for defenders — turning the lens on yourself

The most valuable application of today’s skills: periodically OSINT your own organization to find what attackers will find. Run the entire toolkit on your own assets quarterly. Treat the findings as work items in the harden-the-attack-surface roadmap.

The CISO-level question

“If an attacker spent 4 hours OSINTing us, what would they know that we wish they didn’t?”

That is the question OSINT-for-defenders answers. The output of a quarterly self-OSINT exercise is a prioritized list of exposures:

  1. Forgotten subdomains discovered in CT logs.
  2. Missing or weak SPF/DMARC posture.
  3. Stale documents on the website with leaked metadata.
  4. GitHub repositories with committed secrets.
  5. Employee LinkedIn signals that reveal sensitive technology choices.
  6. Wayback Machine archives of content that should no longer be public.
  7. Shodan-indexed services that should not be internet-facing.

Each becomes a remediation ticket.


12 · The defender’s quarterly OSINT cadence

A practical quarterly routine (4-6 hours per quarter for a mid-size organization):

  1. Website inventory (60 min): browse the public site as an attacker. List every employee name, email, technology mention, vendor mention, project name, and public document.
  2. DNS inventory (45 min): for every domain you own, query A/MX/TXT/SPF/DMARC. Check for missing SPF, weak DMARC, stale NS records.
  3. Certificate transparency (30 min): search crt.sh for every domain. List every subdomain anyone has ever requested a cert for. Investigate unknowns.
  4. GitHub exposure (60 min): search GitHub for your domain, internal hostnames, common secret patterns. Triage findings.
  5. Document metadata audit (45 min): sample 5-10 of your most recently-published PDFs. Run exiftool. Identify leaks. Strengthen the pipeline.
  6. LinkedIn / social signals (30 min): browse your employees’ public LinkedIn. Identify over-shared technology details (e.g., “I administer our Cisco ASA” in a public profile).
  7. Wayback Machine review (30 min): check what historical content is archived for your domains. Identify anything sensitive.
  8. Compile findings + open tickets (30 min): triage findings by severity; open remediation tickets.

Total: ~5.5 hours/quarter. Compared to the cost of an incident, this is the highest-leverage defensive activity most organizations can adopt.


13 · How OSINT feeds the rest of Module 4

L26’s output is the dossier that drives everything later in the module:

  • L27 (Nmap): the IPs and subdomains from L26’s DNS investigation are the targets of L27’s port scans.
  • L28 (Vuln Assessment): the software versions discovered via metadata (and later via Nmap NSE) are the input to vulnerability scanners.
  • L29 (Exploitation): the technologies identified in L26 + L27 + L28 narrow exploit selection.
  • L30 (Web App Security): the public web apps catalogued in L26 are the targets of L30’s web attacks.
  • L31 (Social Engineering): the email-format pattern + employee names + internal project names from L26 are the input to phishing pretexting (defensive tabletop only — no real phishing infrastructure per CSRP convention).
  • L32 (Password Cracking): username conventions inferred from L26’s email pattern feed credential-stuffing and password-spray operations against pre-hashed lab targets.

This is the sequence every offensive engagement follows. By understanding it from the attacker’s perspective, defenders learn where each defensive control matters most.


14 · Cross-references and further reading

CSRP cross-references

  • L17 Mail Security: the SPF/DKIM/DMARC TXT records that today’s cohort queries as OSINT are what defenders publish.
  • L19 Regulation & Compliance: quarterly external-attack-surface-management (EASM) is expected by most compliance frameworks.
  • L20 Cloud Security: Shodan and CT logs frequently catch forgotten cloud resources.
  • L25 Kill Chain: today is Phase 1 in operational detail.
  • L27 (next): active recon — Nmap — built on today’s passive dossier.
  • L31 (later): OSINT-derived pretexting feeds social engineering tabletop.

MITRE ATT&CK mapping

L26’s techniques map to MITRE ATT&CK’s Reconnaissance tactic (TA0043): – T1591 — Gather Victim Org Information (sub-techniques: T1591.001 Determine Physical Locations, T1591.002 Business Relationships, T1591.003 Identify Business Tempo, T1591.004 Identify Roles) – T1592 — Gather Victim Host Information (T1592.001 Hardware, T1592.002 Software, T1592.003 Firmware, T1592.004 Client Configurations) – T1593 — Search Open Websites/Domains (T1593.001 Social Media, T1593.002 Search Engines, T1593.003 Code Repositories) – T1594 — Search Victim-Owned Websites – T1596 — Search Open Technical Databases (T1596.001 DNS/Passive DNS, T1596.002 WHOIS, T1596.003 Digital Certificates, T1596.005 Scan Databases)

Further reading

  • Open Source Intelligence Techniques — Michael Bazzell (regularly updated; the practitioner’s reference).
  • OSINT Framework — osintframework.com — categorized inventory of OSINT tools.
  • The Hacker Playbook 3 — Peter Kim (Chapter 2 on recon).
  • Black Hat Python — Justin Seitz / Tim Arnold (programmatic recon).
  • NSA Cybersecurity Advisory: Mitigating Cloud Vulnerabilities — relevant to OSINT-derived cloud exposure.

Online practice

  • tryhackme.com — “OSINT” and “Sakura Room” rooms.
  • hackthebox.com — “Starting Point” challenges include OSINT components.
  • osintdojo.com — structured OSINT training.
  • (Authorized practice only. No real targets without permission.)

Closing note — the discipline of “defensive intent”

Every technique in this textbook can be used by an attacker against an organization. Every technique can also be used by that organization’s defenders to shrink its public exposure before the attacker arrives. The CSRP graduate’s posture is defensive intent: every offensive skill you acquire becomes a defensive insight.

The discipline is moral and operational. Morally: you commit to the ethical-use contract. Operationally: you run every offensive technique you learn on systems you own or have authorization to test — and you report findings as remediation work, not as attack capability.

L27 introduces the active companion. The contract intensifies. Be patient with yourself, and disciplined with the tools.