Pinned post
Josselin Feist
Aave@Montyly
1.1K Following 5.2K Followers 1.6K Posts
Latest posts 5
There are only a few days left to donate to Echidna
It needs $12k to qualify for the DAO fund and secure its development through 2027
If you use Echidna, share this within your company, tag your boss, and help get it across the finish line :)
It needs $12k to qualify for the DAO fund and secure its development through 2027
If you use Echidna, share this within your company, tag your boss, and help get it across the finish line :)
If Echidna has ever helped you find a bug, now is your chance to help Echidna: initiatives.thedao.fund/initiative/fun…View quoted post on X
Only $12k is needed for it to qualify for the DAO vote
If Echidna has ever helped you find a bug, now is your chance to help Echidna: initiatives.thedao.fund/initiative/fun…
Only $12k is needed for it to qualify for the DAO vote
Only $12k is needed for it to qualify for the DAO vote
After the week we’ve just had, DeFi and blockchain protocols need internal security more than ever
This was just a preview. We are going to see a lot more vulnerabilities and incidents in the coming months
Time to build your in-house security expertise
This was just a preview. We are going to see a lot more vulnerabilities and incidents in the coming months
Time to build your in-house security expertise
Most protocols spend a lot on audits and bug bounties but have zero internal securityView quoted post on X
Launching whohassecurity.com to highlight the ones that do
Having an internal security team should be in every protocol's New Year's resolutions for 2026
Tldr: if you want to find bugs in your codebase, use the latest frontier models
You don’t need a security-specific model outside of a few cases, like doing offensive work (building exploits), working on something sensitive, or running a business that depends on cheap inference
You don’t need a security-specific model outside of a few cases, like doing offensive work (building exploits), working on something sensitive, or running a business that depends on cheap inference
At HackerOne, we benchmarked 18 frontier LLMs against 62 fairly complex vulnerabilities
See which model stands where
(link in the thread)View quoted post on X
