CSRP — Cyber Security Responder & Practitioner
Lesson 26: OSINT and Public Intelligence Gathering
Teacher Lesson Guide
Cyber School · BDO Ziv Haft
CSRP LESSON 26 — TEACHER LESSON GUIDE
Module 4 — Offensive Cyber & Vulnerability Analysis · Lesson 2 of 8 · Course Lesson 26 of 40 Title: OSINT and Public Intelligence Gathering Duration: 4 hours Audience: Cyber Practitioner cohort, mid-Module-4, advancing toward Cyber Analyst role Prerequisite: L25 complete ([Kill-Chain-Mapper] badge), M4 ethical-use briefing acknowledged
1 · Why this lesson exists
L25 introduced the Kill Chain at the framework level. L26 is the first lesson that drills into Phase 1 (Reconnaissance) in operational detail — and it is deliberately the SAFEST lesson in Module 4 to drill into, because nothing students do today produces network traffic visible to a real target. OSINT is the practitioner skill that lives at the intersection of attacker tradecraft and defender hygiene: attackers run it before every targeted operation; defenders run it on themselves to discover what attackers will find.
The lesson exists because most cyber incidents in 2025-2026 are not “zero-day novel attacks.” They start with attackers reading an employee’s LinkedIn, finding a forgotten subdomain in certificate transparency logs, harvesting an internal hostname from a leaked PDF’s metadata, or discovering an exposed API key in a public GitHub commit. The technical work is mundane; the impact is severe. Every cohort member needs the muscle memory to (a) run the OSINT toolkit fluently, (b) recognize the passive/active boundary as a legal-and-ethical line, and (c) flip the lens to audit their own organization’s exposure quarterly.
L26 also reinforces — for the second M4 lesson in a row — the ethical-use contract. Today is the lesson where it would be most tempting to “just try it on a real domain.” That temptation is the exact moment the contract matters. Students leave today knowing they may OSINT only authorized targets in lab environments, and that the same techniques applied to non-authorized targets remain inside legal limits in most jurisdictions only when strictly passive and frequently still violate platform terms-of-service.
Outcome contract: by the end of L26, every learner can (a) define OSINT and articulate the passive-vs-active boundary in legal terms, (b) build a basic OSINT profile of a fictional target from a public website, (c) query A / MX / TXT / SPF / DMARC records and Whois data using dig, nslookup, and whois, (d) extract document metadata with exiftool and identify which fields are sensitive, (e) apply the defender’s OSINT checklist to reduce an organization’s public attack surface, and (f) assemble a 4-element flag from cross-source public clues.
2 · Learning objectives (verbatim)
- Define OSINT and explain the distinction between passive and active reconnaissance — including the legal boundary in major jurisdictions.
- Operate within the legal-and-ethical boundaries of recon (authorized targets, lab environments, no system probing without written permission).
- Use a fictional company website (northwind-academy.local) to build a basic OSINT profile — employee names, email patterns, technology stack, public documents.
- Use Linux DNS/Whois tools (dig, nslookup, whois) to investigate a domain’s public-facing infrastructure including A / MX / TXT / SPF / DMARC.
- Use exiftool to extract metadata from PDFs / DOCX / images and identify sensitive data leakage.
- Apply OSINT-for-defenders thinking — reduce the organization’s public attack surface through quarterly self-OSINT.
3 · Pre-class checklist (T-24h)
| # | Action | Verification |
| 1 | Confirm Oxy lab L26-osint healthy across all per-seat instances | Health endpoint 200 across N seats |
| 2 | Fictional website northwind-academy.local serves cleanly inside the lab (browser-accessible) | curl -I http://northwind-academy.local/ returns 200 |
| 3 | Local DNS simulation answers for northwind-academy.local (A, MX, TXT, SPF, DMARC, hidden subdomain project-aurora.northwind-academy.local) | dig +short northwind-academy.local ANY returns expected fixture |
| 4 | Linux workstation has dig, nslookup, whois, exiftool installed and on PATH | which dig nslookup whois exiftool all return paths |
| 5 | Sample documents pre-staged at /labs/L26/samples/ (≥10 files: PDFs, DOCX, JPGs) including the seeded “INTERNAL: Aurora Project Plan” metadata tag | ls /labs/L26/samples/ ≥ 10 files |
| 6 | Per-instance flag generated and bound to student session ID; validator can score it | Test with known good submission |
| 7 | Verify L25 badge gating ([Kill-Chain-Mapper]) — only L25-pass students enrolled | LMS report |
| 8 | M4 ethical-use disclaimer slide ready (mandatory re-emphasis even though L25 introduced it) | Slide present in deck |
| 9 | Worksheet template printed/digital — the OSINT profile dossier the cohort fills out in Lab 26.1 | Worksheet accessible |
| 10 | Tiered hint timing wired (T=8 concept, T=15 tool, T=20 near-spoiler) | Validator log shows hint events |
Hard stop conditions: if the local DNS simulation answers are missing, Lab 26.2 cannot run as designed — fall back to a documentation walk-through using a captured dig transcript from staging. Do not let learners point dig at real internet domains; the lesson convention is fictional targets only.
4 · Time budget (4 hours)
| Block | Time | Activity | Energy |
| 0:00–0:15 | 15 min | L25 recap + L26 framing + ethical-use re-emphasis | High |
| 0:15–0:30 | 15 min | Quick-win — Extract ONE Recon Finding from a screenshot | High |
| 0:30–1:10 | 40 min | Ch.1 What is OSINT + Ch.2 Passive vs Active | High |
| 1:10–1:20 | 10 min | Break | — |
| 1:20–2:10 | 50 min | Lab 26.1 — Passive Recon on a Fictional Company | Working |
| 2:10–2:20 | 10 min | Break | — |
| 2:20–2:55 | 35 min | Ch.3 DNS as OSINT + Lab 26.2 — DNS and Whois Investigation | Working |
| 2:55–3:25 | 30 min | Ch.4 Metadata as OSINT + Lab 26.3 — Metadata Exposure | Working |
| 3:25–3:30 | 5 min | Ch.5 OSINT for Defenders — reduce public attack surface | Medium |
| 3:30–3:55 | 25 min | Challenge — Public Clues | Independent |
| 3:55–4:00 | 5 min | Wrap + L27 (Nmap) preview | Closing |
5 · Teaching beats — by block
Block 0:00–0:15 · L25 recap + L26 framing + ethical re-emphasis
Opening line: > “Yesterday: you mapped the 7 phases of the Kill Chain. Today: we live inside Phase 1 — Reconnaissance. We learn what an attacker can discover about your organization WITHOUT ever sending a single packet at your systems. And then we flip the lens: we run the same toolkit on ourselves, quarterly, to shrink what attackers can find.”
Mandatory ethical-use disclaimer (verbatim, ~45 seconds — even shorter than L25’s, but mandatory): > “Re-emphasizing the M4 contract: today’s techniques are taught for defensive purposes. We use them on authorized lab targets only. The fictional company today is northwind-academy.local — it does not exist. You do NOT run dig, whois, or exiftool against any real organization during this lesson. Outside this room and outside written authorization, even passive recon can violate platform terms-of-service and, in some jurisdictions, active probing — which we do not do today — can be a crime. Misuse is illegal under CFAA, Computer Misuse Act, Israel’s Computer Law. When in doubt, ask before you act.”
Walk the L26 position in M4: lesson 2 of 8. Preview the day’s arc.
Block 0:15–0:30 · Quick-win — Extract ONE Recon Finding (5-7 min facilitation)
On screen: a screenshot of a fictional company job posting. Sample posting reads: > “Senior SOC Analyst wanted. Tier-2 role. 5+ years experience. Strong skills in Suricata IDS, AWS GuardDuty, Splunk, Linux endpoints. Familiarity with Cisco ASA preferred. Remote, HQ in Tel Aviv.”
Cohort calls out OSINT signals. Anticipated harvest: – Tech stack: Suricata, AWS, Splunk, Cisco ASA → tells attacker what defenses to expect and what evasion to plan – Cloud provider: AWS → narrows attack surface enumeration (S3 buckets, IAM, etc.) – Mail/role hint: “SOC Analyst” + “HQ in Tel Aviv” → location and SOC sophistication implied – Maturity: “Tier-2” implies a tiered SOC exists → larger org
Debrief: ONE job posting just leaked five infrastructure facts. Multiply by 10 postings + 50 LinkedIn profiles + 100 employee bios — the dossier writes itself.
Block 0:30–1:10 · Ch.1 + Ch.2 lecture (40 min)
Ch.1 (20 min) — What is OSINT. Slow down. Define OSINT crisply. Walk the inventory of public sources (websites, LinkedIn, DNS, certificate transparency, metadata, search engines, GitHub, social media, archives, threat intel). Be EXPLICIT about what is NOT OSINT: active probing, login attempts, anything requiring authentication, anything past the public-vs-private line.
Use the rhetorical device: “If a journalist could obtain this without breaking the law, it’s OSINT. If a journalist would need a court order or a leaker, it’s not.”
Ch.2 (20 min) — Passive vs Active. Draw the line clearly. Passive = the target never sees you (you query intermediaries — search engines, third-party Whois services, public archives, certificate-transparency aggregators). Active = the target sees your packets (port scans, DNS zone transfers, banner grabbing).
Common misconception: “Active means logging in.” Correct: active means generating any traffic visible to the target’s logs/IDS. A single TCP SYN packet to a target you don’t have permission to probe is active recon.
Legal nuance: the CSRP rule is “today is passive only.” In some jurisdictions (US under CFAA, Israel under the Computer Law) even a benign port scan against a non-authorized target can constitute unauthorized access. We do not test this boundary on real systems.
The grey zone (~5 min): mention CT log queries at scale (operator-observable), Wayback Machine archiving (creating a new snapshot is closer to active), search engine query logging. Default posture: stay clearly passive; limit to authorized targets.
Block 1:20–2:10 · Lab 26.1 — Passive Recon on a Fictional Company (50 min)
The anchor lab. Per syllabus tasks (verbatim): open northwind-academy.local → identify employee names → identify email-format pattern → identify technology mentions → identify public documents → build a basic OSINT profile.
Watch for: – Learners who skim the site and miss the staff bio pages → coach: “every employee bio page is gold; check the deepest navigation paths.” – Learners who note one email and stop → push for the PATTERN. Is it first.last@? first_last@? firstinitial-last@? The pattern is what an attacker uses to enumerate everyone else. – Learners who list “Suricata, AWS” but don’t write WHY it matters → reinforce the “so what?” column on the worksheet. – Learners who finish in 30 minutes → push them onto the stretch task: cross-reference the staff bios with a hypothetical LinkedIn enumeration (mocked in the lab via additional bio pages).
The “aha” engineered: when a learner finds the staff bio page that lists the email format AND mentions “Project Aurora” — they realize a single page leaked two of the four challenge clues. Don’t tell them this directly; let the realization land in Ch.6.
Block 2:20–2:55 · Ch.3 + Lab 26.2 — DNS as OSINT (35 min)
Ch.3 (10 min) — DNS reveals everything. Walk the record types: A (where the service lives), MX (mail vendor), TXT (SPF/DMARC/DKIM — security posture), CNAME (subdomain dependencies and SaaS vendors), NS (authoritative DNS). Mention certificate transparency briefly as the “every subdomain you’ve ever requested a cert for” reveal — implication: assume every cert subdomain is known to attackers.
Lab 26.2 (25 min) — per syllabus tasks (verbatim): query A records → query MX records → query TXT records → identify SPF/DMARC → review Whois → document exposed services.
Commands the cohort will run (preview on screen):
dig northwind-academy.local A
dig northwind-academy.local MX
dig northwind-academy.local TXT
dig +short northwind-academy.local ANY
nslookup -type=MX northwind-academy.local
whois northwind-academy.local
Watch for: – Learners who don’t know to use +short for cleaner output → quick tip mid-lab. – Learners who confuse dig and nslookup outputs → coach: dig is more flexible for production work; nslookup is fine for quick checks. – Learners who note “SPF exists” without parsing the content → push for what’s actually in the SPF record (v=spf1 include:_spf.northwind-academy.local -all vs. ~all vs. ?all — the -all vs. ~all distinction matters defensively). – Learners who find project-aurora.northwind-academy.local and ignore it → the hidden subdomain is a challenge clue. Don’t telegraph; let them notice.
Block 2:55–3:25 · Ch.4 + Lab 26.3 — Metadata as OSINT (30 min)
Ch.4 (10 min) — The forgotten leak. Walk what metadata typically reveals in PDF / DOCX / JPG. Mention famous cases (declassified PDFs revealing redacted authors, DOCX revision history leaks in M&A, JPG EXIF GPS exposing home addresses). Defender’s response: strip metadata before publication.
Lab 26.3 (20 min) — per syllabus tasks (verbatim): open provided files → run exiftool → identify metadata fields → mark sensitive metadata → recommend metadata-removal controls → write a short exposure report.
Commands the cohort will run:
cd /labs/L26/samples/
exiftool *.pdf
exiftool *.docx
exiftool *.jpg
exiftool -a -G1 sample01.pdf # -G1 groups by metadata family
exiftool -all= -overwrite_original copy_of_sample.pdf # strip metadata
Watch for: – Learners who see “Author: J. Smith” and shrug → push: “If J. Smith is the marketing manager, you’ve just learned a target. If J. Smith is the CISO, you’ve just learned the author of an ‘official’ document.” – Learners who find the “INTERNAL: Aurora Project Plan” tag and don’t flag it as sensitive → coach: “Internal-only” tags should NEVER appear in published documents. – Learners who recommend “just be careful” → push for SPECIFIC controls: automated metadata-stripping pipeline, Office “Document Inspector,” qpdf –linearize –object-streams=generate, exiftool -all= pre-publish hook.
Block 3:25–3:30 · Ch.5 — OSINT for Defenders (5 min)
Brief framing lecture. The CISO-level question: “If an attacker spent 4 hours OSINTing us, what would they know that we wish they didn’t?” Walk the defender’s quarterly checklist (website inventory, DNS inventory, CT log search, GitHub exposure, document metadata, LinkedIn signals, Wayback Machine). Close with: “You now have the toolkit. Run it on yourself quarterly. The findings drive the harden-the-attack-surface roadmap.”
Block 3:30–3:55 · Challenge — Public Clues (25 min)
Hand-off: > “25 minutes. Independent work. Browser + Linux terminal. Find FOUR clues: email format on the website, hidden subdomain in DNS, sensitive metadata tag in a PDF, and a hidden project name that appears in all three. Combine per the validator’s instructions. Submit ULC{…}. Three tiered hints unlock at T=8 (concept), T=15 (tool), T=20 (near-spoiler). Go.”
Walk the room. Don’t help unless asked. Mid-challenge intervention scripts: – Stuck at minute 5: “Where would you look first if you were the attacker?” – Stuck at minute 12 (hint T=8 already): “Which Linux tool gives you DNS records?” – Stuck at minute 18 (hint T=15 already): “What’s the hidden project name? It appears more than once.”
Block 3:55–4:00 · Wrap + L27 preview (5 min)
Quick recap: M4 progress 2/8. Award [Passive-Recon-Analyst] to Challenge-pass learners. Preview L27 (Nmap): “Today: passive recon — target never sees you. Next time: ACTIVE scanning — Nmap generates traffic the target sees. We move from ‘what is public’ to ‘what is reachable.’ Same ethical contract applies, intensified.”
6 · Differentiation
| Learner | Symptom | Move |
| Has bug-bounty / OSINT background | Restless during Ch.1/Ch.2 | Stretch task: enumerate CT logs for the fictional domain via the staged crt.sh-mirror in the lab; present findings to cohort |
| Defender-only background | Hesitates to “investigate” the fictional company | Reframe: “You’re auditing your own org’s exposure. The fictional target is YOUR org.” |
| Linux-shell-shaky | Stuck on dig / exiftool syntax | Pair with shell-fluent peer; provide the cheat-card from the workbook; the lesson is recon, not bash |
| Hebrew-first | Asks for HE term equivalents | HE OSINT glossary in workbook; cyberPedia entries for “passive recon,” “metadata,” etc. |
| Ethics-anxious (“can I get in trouble for any of this?”) | Reluctant to run dig | Re-emphasize: fictional target, lab DNS, no real internet traffic generated. Show the lab network diagram |
| Wants to run it on a real company “just to see” | Curious but boundary-testing | Firm no. Explain: even passive techniques like Whois can be rate-limited and have terms-of-service. Real OSINT happens under written authorization in real engagements |
7 · Common pitfalls + recovery scripts
Pitfall 1: “OSINT isn’t really hacking; why are we spending a whole lesson?” Recovery: “Because most attacks START here. Reading public sources costs the attacker nothing, leaks no traffic, and produces the target dossier that drives every later phase. Defenders who skip this lesson don’t know what attackers see — they harden the wrong things.”
Pitfall 2: “Isn’t passive recon legal? Why so careful?” Recovery: “Passive recon is generally legal — but ‘generally’ isn’t ‘always.’ Whois has terms of service. LinkedIn scraping has been litigated. CT log queries at scale are observable. And the line between passive and active is thinner than it looks. CSRP convention: authorized targets only, even for passive techniques, because the habit is what matters.”
Pitfall 3: “exiftool just dumps a wall of text — what’s important?” Recovery: “Three families: identity fields (Author, Last-Modified-By, Creator), system fields (CreatorTool, Producer, file paths), and content fields (comments, tags, GPS). For each finding, ask: ‘If a journalist published this, would the source be embarrassed?’”
Pitfall 4: “I found the subdomain but I don’t know how it’s a ‘clue.’” Recovery: “Look at the name. Subdomain names are not random — they reflect the project they serve. A name leaks the project’s existence.”
Pitfall 5: “My organization doesn’t publish PDFs / DOCX externally — does this lesson apply?” Recovery: “Almost every org does. Marketing pages, white papers, annual reports, regulatory filings, job descriptions. If you publish anything in a Word format, the metadata pipeline applies.”
Pitfall 6: “I want to try this on a real domain right after class.” Recovery: “If it’s your own organization’s domain, with manager approval, that’s the defender’s quarterly self-OSINT — that’s exactly the application we want. Anything else: written authorization first. The CSRP graduation contract is non-negotiable on this.”
8 · Materials checklist
- Gamma deck open with M4 ethical-use disclaimer slide visible
- Textbook PDF accessible to cohort
- Workbook digital/printed (OSINT profile worksheet inside)
- Oxy lab L26-osint healthy across all per-seat instances
- Fictional northwind-academy.local website serves with all staff bio pages, public documents, and the hidden Aurora references
- Local DNS fixtures answer for A / MX / TXT / SPF / DMARC / Whois / project-aurora subdomain
- Sample documents (≥10) at /labs/L26/samples/ — at least one PDF with “INTERNAL: Aurora Project Plan” metadata tag
- Linux workstation tooling verified (dig, nslookup, whois, exiftool on PATH)
- Slack #csrp-l26 channel open for cohort questions
- Visible timer for Challenge (25-min countdown)
- M4 ethical-use disclaimer slide confirmed in deck
- Validator + flag-gen script tested with one sample student session ID
- Tiered hint events wired (T=8, T=15, T=20) and verified in test run
9 · Assessment & badging
Per-lesson quiz (LearnDash): 8 questions sampled from the L26 Question Bank, 75% pass. Labs 26.1 / 26.2 / 26.3: progress points; rubric checkpoints scored by the Oxy validator (see file 07). Challenge — Public Clues: required for badge. Submit ULC{…} flag derived per the validator’s instructions, assembled from the 4 cross-source clues. Badge: [Passive-Recon-Analyst] — auto-issued on Challenge pass.
10 · Cross-references
- L17 Mail Security: SPF / DKIM / DMARC TXT records are what defenders publish; today’s cohort queries them as OSINT
- L19 Regulation & Compliance: quarterly self-OSINT is a control most compliance frameworks expect under “external attack surface management”
- L20 Cloud Security: Shodan and certificate transparency frequently catch forgotten cloud resources
- L25 Kill Chain: today is Phase 1 (Reconnaissance) — the foundation under every later phase
- Forward to L27 (Nmap): the active companion — the same target dossier built today drives where Nmap scans tomorrow
- Forward to L31 (Social Engineering): OSINT-derived email patterns and employee names are the input to phishing pretexting (taught defensively, lab-only)
- Forward to L33+ (Module 5 SOC): “external attack surface management” is a daily SOC discipline; today’s checklist is the operational basis
11 · Sign-off
| Role | Name | Date |
| Lesson author | ||
| BDO Ziv Haft reviewer | ||
| Pedagogy reviewer | ||
| Cyber technical reviewer | ||
| Ethical-use disclaimer approval | ||
| Approved for cohort delivery |