Trail of Bits

Hash several inputs by concatenation and the boundaries vanish: different sequences give the same SHA-256 digest.

Trail of Bits' SequenceHash appends each input's length as a fixed 128-bit integer, so every sequence encodes one way. Rust, Go and Python.

blog.trailofbits.com/2026/10/02/seq…

Credit to Trail of Bits for showing where agents pay off before a security review even starts.

For one audit they spent six months having agents build the tooling first: an LSP server, a decompiler, a static analysis engine and a Lean model of the VM. The analyses flagged 400+ places where type validation could be tightened. The Lean work produced 95 machine-checked correctness proofs.

Their line: "Today, a failed side project only costs tokens."

blog.trailofbits.com/2026/09/18/aud…

We're excited to announce that @trailofbits, an independent security research firm, has completed a security review of Kamui's RWA infrastructure.

We’re bringing together some incredible teams across the Arbitrum ecosystem to support founders building at Founder House Singapore.

> @Dune
> @Paxos
> @ZeroDev
> @GMX_IO
> @trailofbits
> @pendle_fi
> @Quicknode
> @RobinhoodCrypto
> SBI Digital Markets

Innovating across tokenization, AI and programmable finance?

Join us and Robinhood Chain in Singapore from October 23rd-25th.

Register here:

luma.com/openhouse-sing…

@ShieldFi is a fortress. Over the past years we've done:

- A Lean impl of the proof system (thx @Zcash for paving the way!)
- ACL2 impl of our circuits and consensus (thx Kestrel!)
- TLA+ impls for smart contract invariants (thx @trailofbits !)
- An easycrypt impl for the VM

We're hiring a Solutions Architect to own the technical strategy for our key clients, stay with accounts after each engagement, and design the program that follows.

Remote, US or Canada. apply.workable.com/trailofbits/j/…

Thank you to @trailofbits for sponsoring OAIC 2026! They recently published research where they gave GPT 5.6-Cyber one task: escape a QEMU/KVM VM used to sandbox agents. In its final escape, the agent chained three 0-days into a working exploit.

blog.trailofbits.com/2026/08/26/vms…

We're sponsoring @arbitrum's Open House Singapore, hosted in collaboration with Robinhood Chain, to support founders across both the online Buildathon and Founder House.

From October 23-25, Founder House Singapore brings selected teams together for 3 days of product, technical, and go-to-market support to help turn early ideas into launch-ready onchain products, while competing for a share of $300K in prizes and grants.

Founder House teams will get a remote mentoring session with @gintaria from our team.

Apply here: luma.com/openhouse-sing…

Hashing multiple values together is easy to get wrong, and mistakes can lead to forgeries worth millions in zero-knowledge proofs. We built SequenceHash and SequenceMAC to fix that for any hash function, not just Keccak. blog.trailofbits.com/2026/10/02/seq…

September Trail of Bits Tribune: Fact-checking 1Password's AI patching benchmark, and a signature-forgery bug in Miden's zkVM. Also: a string-slicing bug in Lean, 27 advisory submissions to OpenClaw, and seven new security reviews.

mailchi.mp/trailofbits/se…

We've reviewed AvalancheGo, libevm, and Subnet EVM for @AvaLabs. We also built Slither, Echidna, and Medusa, and published 391 blockchain security reviews.

Avalanche builders can now request an audit from us through the Avalanche Audit Marketplace. build.avax.network/audits/firms

Threshold signatures secure billions in crypto and increasingly run inside TEEs. But MPC and TEEs make different bets on trust: MPC spreads it across independent parties; TEEs concentrate it in the hardware manufacturer and its attestation infrastructure. blog.trailofbits.com/2026/09/25/don…

We're the only red team for @ARIA_research's Safeguarded AI programme, led by @AmmannNora.

Each development cycle, we attack the code, its proofs, and everything those proofs rest on: the models, specs, assumptions, and deployment choices behind every team's assurance case.

We're hiring a Senior Developer Relations Engineer who builds and ships, with the security depth to work alongside our engineers and open-source maintainers.

Apply here: apply.workable.com/trailofbits/j/…

We reviewed @bee__computer Private Compute's mobile apps, servers, and AMI builder and reported 21 issues. Amazon fixed all 7 high-severity findings and published the full report.
github.com/trailofbits/pu…

Our @openclaw security assessment produced 27 advisories, 3 hardening PRs, and 1 architectural submission.

Read the report:
github.com/trailofbits/pu…

A 2012 USENIX paper brought XML Signature Wrapping (XSW) attacks against SAML into the mainstream. SAML has repeatedly shown that complexity is the antithesis of security. Matt Schwager names 5 protocol design flaws enabling SAML's insecurity: blog.trailofbits.com/2026/09/21/sam…

We audited parts of @0xMiden's zkVM, written in a custom assembly language called MASM. MASM originally had no developer tooling, so we used Claude and Codex to build a decompiler and static-analysis engine. It flagged a signature-forgery bug that could've cost millions. blog.trailofbits.com/2026/09/18/aud… @fegge

We’re hiring an Engineering Director to lead Application Security. We want someone who still finds bugs in code, builds excellent teams, and has a clear view of how AI impacts security audits.

Apply here: apply.workable.com/trailofbits/j/…

.@1Password's FLAWED report says AI models produce a clean security fix only 26% of the time.

Defenders shouldn't take that number seriously.

• The six vulnerabilities were handpicked because their fixes were complex. Clean-fix rates ran from 3% to 60% depending on the bug, and the report averaged them together.

• Agents set up to fail were counted in the headline figure. Two of 1Password's prompts instructed the agent to apply the wrong fix. Those trials make up 22% of the data. One evaluation mode prevented the agent from compiling or running any code, and it accounts for 36% of the data.

• The report ran two models, GPT-5.5 at medium effort and Opus 4.8 at high. Neither was tested at its highest available setting, so the report says nothing about how more effort or stronger models change the results.

• Several instruction and grading errors further undercut the headline, and are elaborated upon in the attached blog.

We've spent four months submitting hundreds of AI-authored patches to widely adopted open-source projects as part of Patch the Planet.

Our experience didn't match 1Password's report, so we did a full analysis across 186 AI-authored pull requests and 33,500 subsequent commits, benchmarked against 2,265 human-authored patches we graded across years of security engagements. blog.trailofbits.com/2026/09/15/1pa…