How to Intercept HTTP/3 Traffic: QUIC and Qpack
anof-cyber
68 points
8 comments
November 16, 2026
Related Discussions
Found 5 related stories in 174.5ms across 13,334 title embeddings via pgvector HNSW
- Qpack amplification attacks against HTTP/3 servers kozika · 23 pts · October 21, 2026 · 70% similar
- HTTP/3 is faster with some work and help from Cloudflare saikatsg · 13 pts · October 31, 2026 · 52% similar
- Prompt injection in HTTP headers: what 217M web responses show panoramas4good · 84 pts · October 15, 2026 · 46% similar
- Show HN: Q3Edit – Edit and play Quake 3 maps in the browser drdator · 79 pts · July 18, 2026 · 45% similar
- I captured 72 hours of idle Android packets behind pfSense youngmanyk · 42 pts · September 19, 2026 · 44% similar
Discussion Highlights (2 comments)
rbartelme
The key requirement here is the client has to trust the proxy's CA. It's a client-side MITM, and the article itself notes pinned apps reject it. QUIC's crypto is fine. HTTP/3 just stopped being a blind spot for corporate TLS-inspection boxes. It's more interesting to split tunnel designs by where the trust lives: - ngrok / Cloudflare Tunnel: the edge terminates TLS by design, and your public hostname is ordinary WebPKI. Anyone who controls a visitor's trust store (inspection appliance, malware root cert) can do exactly this, and the provider can see plaintext anyway. - iroh-style P2P QUIC: there's no CA in the loop. You dial a public key (the EndpointId), and the TLS handshake authenticates raw keys, so it's pinned by construction. Relays forward ciphertext they can't terminate. A rogue root cert gets an attacker nothing; they'd need the peer's secret key. So "can someone MITM my HTTP/3?" really comes down to "who can add to the set of keys my client trusts?" For WebPKI that's a long list. For key-addressed transports it's nobody: the trusted key is the address you dialed, unless an endpoint's secret key (or the machine itself) is already compromised.
opem
Stupidity over a vibecoded slop software?