Agent-Killers
Popular repositories Loading
-
Agent-Killers
Agent-Killers PublicIt is an evidence-first benchmark arena for coding agents and human engineers, designed to test not only whether an agent can produce a correct patch, but whether it can survive hostile instruction…
PowerShell 1
Repositories
- Agent-Killers Public
It is an evidence-first benchmark arena for coding agents and human engineers, designed to test not only whether an agent can produce a correct patch, but whether it can survive hostile instructions, protect scope, pass hidden checks, kill mutations, reproduce its result, and produce evidence that another party can verify.
Top languages
Loading…
Most used topics
Loading…