Enclave
@EnclaveAI
Mistral Large 4 verified 0 of 11 runs in the AI Hacking Race.
In the Grafana and Jenkins challenges we reviewed in depth, it found the planted vulnerability and built a working primitive, but never turned it into verified code execution.
It's a strong auditor but weak at
In the Grafana and Jenkins challenges we reviewed in depth, it found the planted vulnerability and built a working primitive, but never turned it into verified code execution.
It's a strong auditor but weak at
enclave.aiEnclave: AI Hacking Race | Benchmark Leaderboard
0 27