Security Exploits & Vulnerabilities
You cannot build a wall if you do not know how walls get broken. The best defenders are engineers who understand offensive techniques deeply enough to anticipate how an attacker will approach their systems. This section is not about teaching you to hack — it is about teaching you to think like someone who would, so that your defenses are not theoretical but battle-tested.
Every page in this section follows the same structure: how the attack works, then how to defend against it. No exploit is presented without its corresponding mitigation.
Related: Security Overview | OWASP Top 10 | Supply Chain Security
Why Study Exploits?
The Defender's Paradox
An attacker needs to find one vulnerability. A defender needs to prevent all of them. The only way to close the gap is to understand how attackers think, what they look for, and how they chain small weaknesses into devastating breaches.
- Threat modeling requires attack knowledge. You cannot enumerate threats you have never heard of. Studying real exploits builds your threat vocabulary.
- Code review becomes sharper. Once you have seen how a buffer overread leaks private keys (Heartbleed) or how string interpolation enables remote code execution (Log4Shell), you catch similar patterns in your own code.
- Architecture decisions improve. Knowing that containers are not security boundaries by default changes how you design isolation. Understanding supply chain attacks changes how you manage dependencies.
- Incident response needs speed. When a new CVE drops, teams that already understand the attack class can assess impact and deploy mitigations in hours instead of days.
Attack Taxonomy
mindmap
root((Security<br/>Exploits))
Supply Chain
Dependency compromise
Build system injection
Social engineering maintainers
Registry poisoning
Memory Safety
Buffer overflow / overread
Use-after-free
Integer overflow
Format string
Injection
SQL injection
NoSQL injection
Command injection
Template injection SSTI
LDAP injection
Logic & Web
XSS DOM, reflected, stored
CSRF
SSRF
Authentication bypass
Privilege escalation
Cryptographic
TLS protocol attacks
Timing attacks
Nonce reuse
Weak hashing
Infrastructure
Container escapes
Cloud misconfigurations
Kernel exploits
CPU side channels
By Attack Vector
| Category | What Gets Exploited | Example Incidents | Pages |
|---|---|---|---|
| Supply Chain | Dependencies, build systems, maintainer trust | XZ backdoor, SolarWinds, Log4Shell | XZ Backdoor, SolarWinds, Log4Shell |
| Memory Safety | Missing bounds checks, pointer misuse | Heartbleed, Dirty Pipe, Dirty COW | Heartbleed, Dirty Pipe |
| Injection | Untrusted data mixed with code | SQL injection, SSTI, command injection | Advanced Injection |
| Web/Logic | Browser trust model, CSP gaps, DOM manipulation | XSS, mXSS, CSP bypass | Advanced XSS |
| Hardware | CPU speculative execution, microarchitecture | Spectre, Meltdown | Spectre & Meltdown |
| Infrastructure | Container isolation, cloud IAM, metadata services | Container escapes, S3 exposure, SSRF to IMDS | Container Escapes, Cloud Misconfigs |
| Cryptographic | Protocol weaknesses, implementation flaws | BEAST, POODLE, timing attacks | Crypto Attacks |
By Impact
graph LR
subgraph "Impact Severity"
RCE["Remote Code<br/>Execution"]
PRIV["Privilege<br/>Escalation"]
DATA["Data<br/>Exfiltration"]
DOS["Denial of<br/>Service"]
INFO["Information<br/>Disclosure"]
end
RCE --> |Log4Shell, SSTI| EX1[Attacker runs arbitrary code]
PRIV --> |Dirty Pipe, Container Escape| EX2[Attacker gains root/admin]
DATA --> |S3 exposure, SQLi| EX3[Attacker steals sensitive data]
DOS --> |Spectre mitigations, ReDoS| EX4[System unavailable or degraded]
INFO --> |Heartbleed, timing attacks| EX5[Attacker reads secrets from memory]
style RCE fill:#dc2626,color:#fff
style PRIV fill:#ea580c,color:#fff
style DATA fill:#d97706,color:#fff
style DOS fill:#2563eb,color:#fff
style INFO fill:#7c3aed,color:#fff
The Exploit Lifecycle
Every major vulnerability goes through a predictable lifecycle. Understanding this timeline helps you prepare your incident response process.
sequenceDiagram
participant R as Researcher
participant V as Vendor
participant P as Public
participant A as Attackers
R->>R: Discover vulnerability
R->>V: Private disclosure (Day 0)
V->>V: Develop patch
Note over V: 90-day disclosure deadline<br/>(Google Project Zero standard)
V->>P: Publish CVE + patch
P->>P: Media coverage begins
A->>A: Reverse-engineer patch
A->>A: Develop exploit
Note over A: "n-day" exploitation window
A->>P: Mass scanning begins
Note over P: Organizations that have not<br/>patched are now targets
Timeline of Exploitation
| Phase | Typical Duration | What Happens |
|---|---|---|
| 0-day | Unknown to vendor | Only discoverer knows; may be sold on exploit markets ($50K-$2.5M) |
| Responsible disclosure | 90 days (standard) | Vendor develops and tests patch in private |
| Patch release | Day 0 of public knowledge | CVE published, patch available |
| n-day window | 1-30 days | Attackers reverse-engineer patch, develop exploits, mass-scan for unpatched systems |
| Long tail | Months to years | Unpatched legacy systems remain vulnerable indefinitely |
The Patch Gap
The most dangerous period is between patch release and patch deployment. Log4Shell was being mass-exploited within 24 hours of public disclosure. Heartbleed was exploited within hours. Your patching speed is your security posture.
Responsible Disclosure
Responsible (or coordinated) disclosure is the practice of privately reporting vulnerabilities to the vendor before making them public, giving them time to develop and release a fix.
Disclosure Models
| Model | How It Works | Pros | Cons |
|---|---|---|---|
| Coordinated disclosure | Report to vendor, agree on timeline (typically 90 days), publish after patch | Gives vendor time to fix, protects users | Vendor may delay indefinitely |
| Full disclosure | Publish immediately, no vendor notification | Forces immediate action, no suppression | Users are exposed before patch exists |
| Bug bounty | Report through structured program, receive payment | Financial incentive for researchers, structured process | Scope limitations, triage delays |
Bug Bounty Programs
Major bug bounty platforms and notable payouts:
| Platform | Notable Programs | Typical Payouts |
|---|---|---|
| HackerOne | US DoD, GitHub, Shopify, Uber | $500 - $100,000+ |
| Bugcrowd | Tesla, Mastercard, Atlassian | $500 - $50,000+ |
| Google VRP | Chrome, Android, Cloud | Up to $250,000 (Chrome full chain) |
| Microsoft MSRC | Windows, Azure, Office 365 | Up to $250,000 (Hyper-V) |
| Apple Security Research | iOS, macOS | Up to $2,000,000 (zero-click kernel) |
Legal Boundaries
Security research without authorization is illegal in most jurisdictions. Always:
- Only test systems you own or have explicit written permission to test
- Use dedicated lab environments or CTF platforms for practice
- Follow the scope rules of any bug bounty program exactly
- Document your methodology — good-faith research is a legal defense, but only if documented
Section Map
| Page | Attack Class | Year | Impact |
|---|---|---|---|
| XZ Utils Backdoor | Supply chain, social engineering | 2024 | Near-miss: backdoor in core Linux utility |
| Log4Shell | Injection, RCE | 2021 | Billions of devices, internet-wide scanning |
| SolarWinds | Supply chain, nation-state | 2020 | 18,000+ organizations including US government |
| Heartbleed | Memory safety | 2014 | 17% of SSL servers, 2 years undetected |
| Spectre & Meltdown | Hardware, side-channel | 2018 | Every modern CPU affected |
| Dirty Pipe | Kernel exploit, privilege escalation | 2022 | Linux kernel 5.8+ |
| Container Escapes | Infrastructure, isolation bypass | Ongoing | All major container runtimes |
| Advanced Injection | Injection | Ongoing | #3 OWASP Top 10 |
| Advanced XSS | Web, client-side | Ongoing | #7 OWASP Top 10 |
| Cloud Misconfigurations | Infrastructure, IAM | Ongoing | Capital One ($80M fine), countless others |
| Cryptographic Attacks | Cryptography | Ongoing | TLS protocol evolution driven by attacks |
How to Use This Section
- Start with the case studies (XZ Backdoor, Log4Shell, SolarWinds, Heartbleed) — they provide narrative context that makes the technical details memorable.
- Move to attack classes (Injection, XSS, Container Escapes, Cloud Misconfigs) — these generalize the patterns you saw in the case studies.
- Study the hardware and crypto attacks (Spectre/Meltdown, Crypto Attacks) — these challenge your assumptions about what is trustworthy (CPUs, math).
- Cross-reference with defensive sections — every page links back to the relevant OWASP, encryption, zero trust, and supply chain pages.
Building Intuition
Read each case study and ask yourself: "Would I have caught this in code review? Would my CI pipeline have flagged it? Would my monitoring have detected it in production?" If the answer is no, you have found a gap in your defenses.
Further Reading
- Security Overview — threat modeling, security principles, section map
- OWASP Top 10 — the ten most critical web application risks
- Supply Chain Security — SLSA, SBOMs, Sigstore, dependency scanning
- Encryption — cryptographic fundamentals that underpin many defenses
- Zero Trust — assume breach, verify everything