727559c · Security Portfolio

Clip #3: upgrade crypto research to a live forgery demo

Replace the analysis clip with the actual batch-forgery PoC running: an
INVALID Ring-Pedersen proof accepted by the real verifier ~1/256 (gamma=0
collision from the 8-bit type confusion), with a gamma!=0 control proving
the cause. Demonstrates the break instead of just explaining it.

727559c Zion Boggan committed on May 30, 2026 (3 weeks ago)

assets/hero/crypto-demo-poster.jpg +0 -0

Binary file not shown

assets/hero/crypto-demo.mp4 +0 -0

Binary file not shown

index.html +4 -3

		@@ -263,9 +263,10 @@
		<source src="/assets/hero/crypto-demo.mp4" type="video/mp4">
		</video>
		<p class="democap"><span class="step mono">03 · CRYPTO RESEARCH</span> Not a lab - a real flaw I
	-	found and responsibly disclosed in <b>Fireblocks' MPC threshold-signature library</b>: a
	-	one-byte type confusion silently cut a 40-bit security check down to 8 bits
	-	(<b>~4.3 billion× weaker</b>). The difference between running tools and reading the crypto.</p>
	+	found and responsibly disclosed in <b>Fireblocks' MPC threshold-signature library</b>. A
	+	one-byte type confusion cut a 40-bit check down to 8 bits, so I <b>forge an invalid proof the
	+	production verifier accepts</b> - about 1 in 256 tries, with a control run that proves the cause.
	+	The difference between running tools and reading the crypto.</p>
		</div>
		</div></section>