PortSwigger Research Impossible XSS SOLVED (Ep. 183)
Source: YouTube · Critical Thinking - Bug Bounty Podcast · published Jul 16, 2026 · 1:14:47
This episode of Critical Thinking covers AI-powered bug bounty wins — deep link prompt injection, CSPT-to-ATO with a 2FA bypass — plus client-side Sanitizer API research, GitHub agentic workflow injection, and a Razor SSTI blocklist bypass 2:45.
Key Takeaways:
• A mobile app deep link prompt injection used Unicode tag characters to poison AI memory, shimming all future links to an attacker proxy for persistent leakage; a similar chain paid ~$6K 6:10
• AI bounties pay exceptionally well (0.5% premiums, $500 minimums), and AI is good at hacking AI with a proper harness 9:41
• Picking the HackerOne program with fastest triage (16h to bounty) led to three highs paid within an hour — fast response is a motivation "cheat code" 15:27
• A CSPT chain to full ATO bypassed 2FA by sending _proto in the X-2FA-Code header — the inherited prototype property passed the lookup check — for $15K 20:30
• Searchlight Cyber bypassed Chrome's Sanitizer API via SVG <animate> href injection and a KURL parser discrepancy where invalid JavaScript URLs fail open, then normalize into executable URIs 36:00
• Masato Kinugawa solved PortSwigger's "impossible" innerHTML lab using the selectedcontent element; Kevin Mizu adapted it into a universal CSPT gadget for Cloudflare's cdn-cgi endpoint 47:02
• GitHub's agentic workflows were prompt injected via a public issue to exfiltrate private repo contents — the word "additionally" defeated guardrails, with overly broad org-wide token scoping as root cause 54:22
• A Razor SSTI bypassed keyword blocklists using ASCII-built strings, .NET reflection, and the dynamic keyword — a 25-line payload yielding a $2K critical 1:01:56
Core message: creative gadget chaining still requires human artistry — AI assists recon and reporting, but the best-paying exploit chains are assembled by hand.
Sources:
Generate CPE Credits
Generate a professional CPE document from this video's transcript.
Estimated credit: 1 CPE hours
Estimate uses the video runtime (1 hour ≈ 1 CPE, rounded to the nearest 0.5, minimum 0.5, maximum 2.0). The final amount can be lower after review, never higher.
Topic: Cybersecurity. Commonly maps to: Security and Risk Management, Security Operations. Exact CISSP domains are assigned during generation.
Free account. One generation at a time, with a daily limit.
CPEBuddy is independent and not affiliated with or endorsed by ISC2, ISACA, or any certification body. Exports are formatted for common CPE submissions; acceptance is at your certification body's discretion.
Transcript Preview
First 800 characters of the transcript
In this specific scenario, I I said like, "Oh, this is a child safety proxy." Childsafetyproxy.com, you know, right? [laughter] And it's like, you know, if you don't do this, the children the children [music] best part of when you can just, you know, critical thing, right? [music] Yeah, >> dude. >> If you've been in the Bug Bounty Recon game for any period of time, you know how beautiful it is when some unsuspecting dev or DevOps guy spins up what's clearly supposed to be an internal facing server and accidentally just gives you the keys to the kingdom, right? It's a big payday. It's an easy win. It's a beautiful thing, right? And unfortunately for us, Threat Locker also knows about that and that's why they created their zero trust network access product. Okay, check this out. This is a pr…