Understanding Agents: Code Coverage for Coding Agents
When an agent audits your codebase, a common question is: what did it actually read and with what intent? Current tools don't answer that. We built a prototype and open-sourced it.
Security research and insights from the Asymmetric Research team.
When an agent audits your codebase, a common question is: what did it actually read and with what intent? Current tools don't answer that. We built a prototype and open-sourced it.
With support from Solana Foundation, we're launching STRIDE, a comprehensive security program that sets clear standards for ecosystem projects.
The most persistent security misinformation doesn't come from obscure corners of the internet, but from official docs, learning resources, and popular LLMs. Learn about the Solana vulnerabilities that aren't, and why they keep spreading.
When an agent audits your codebase, a common question is: what did it actually read and with what intent? Current tools don't answer that. We built a prototype and open-sourced it.
With support from Solana Foundation, we're launching STRIDE, a comprehensive security program that sets clear standards for ecosystem projects.
The most persistent security misinformation doesn't come from obscure corners of the internet, but from official docs, learning resources, and popular LLMs. Learn about the Solana vulnerabilities that aren't, and why they keep spreading.
With the launch of DoubleZero’s mainnet, we’re grateful to celebrate alongside our Cohort 0 partners. This milestone strengthens
Relay Protocol's contracts trusted Ed25519 verification without validating offsets, opening the door to forged allocator signatures and potential double-spends. Learn about the bug, the risks it posed to cross-chain liquidity, and how the issue was addressed.
A new instruction broke the flash loan logic, creating a way to borrow without repaying and putting $160M at risk. We explain the vulnerability, potential impact, and how it was fixed.
An attacker posing as a well-known web3 founder messaged one of our engineers via Telegram. Rather than ignoring the attempt, we isolated and analyzed the payload in a controlled environment, turning a live phishing attempt into a learning opportunity.
Some of the most devastating vulnerabilities stem from complexity, inconsistency, and chaos. This post explains why predictable, well-formed code is the foundation of security.
In the year 2050, a malformed JSON input lead to the complete shutdown of the Replicant P2P network. Today, we'll reproduce this bug class in ~100 lines of code.
Cross-program invocation (CPI) is the mechanism on Solana through which one program calls another. It's used for system instruction calls, SPL token transfers, custom program execution, and even event emissions, making it a core part of writing functional programs in Solana.
This blogpost details a vulnerability identified in Titan's Helix MEV Relay which could be abused by trusted proposers to perform equivocation attacks, during which private transaction information is prematurely revealed, leading to potential transaction reordering.
In this blog post, we will show how a small difference in SSZ deserialization between the Prysm and Lighthouse clients could have allowed an attacker to severely degrade Ethereum consensus.
Subscribe to be notified whenever we publish new security research.