LABS · EVERY RUN

Every lab run.

Each lab's check, every time it ran, newest first. The rows come from the records this site publishes and nothing else: when it ran, on which runner and machine class, the commit it ran, the record hash and what the check printed. Two runs that computed the same thing from the same commit hold the same record hash, wherever they ran.

119 runs, newest first. Hashes are shortened; each record holds them in full.
RanLabRunnerMachine classCommitRecord hashThe check printedRecord
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavyc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latestc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavyc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latestc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latestc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavyc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latestc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavyc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latestc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavyc169ad9bf419d9215e2bSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latestc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavyc169ad9b9e1467d6b2c8SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyc169ad9bf2c48fc2a28aSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latestd934f799f6b457a9cb81SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavyd934f799f6b457a9cb81SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latestd934f799a964b2f3fc74SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavyd934f799a964b2f3fc74SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestd934f799feb80a48b459SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyd934f799feb80a48b459SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latestd934f799f6b457a9cb81SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavyd934f799f6b457a9cb81SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latestd934f799a964b2f3fc74SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavyd934f799a964b2f3fc74SAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestd934f799feb80a48b459SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyd934f799feb80a48b459SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latest58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavy58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latest58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavy58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latest58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavy58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latest58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavy58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?org-hostedorg-ci-heavy58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavy58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 13: Does an agent company get better, measured blind?github-hostedubuntu-latest58e84a9d9aaa4dcaaff0SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latest58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest88779efe04fa0d5ba8deSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latest58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavy58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavyedc746aaa43b8f02a211SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latestedc746aaa43b8f02a211SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?github-hostedubuntu-latest58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 08: Can an agent fix a red build?org-hostedorg-ci-heavy58e84a9d6a7cf1bd96baSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latestedc746aaa43b8f02a211SAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavyedc746aaa43b8f02a211SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy58e84a9d58eff38051eaSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latest137f98228a49047402daSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavy137f98228a49047402daSAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyccc31e2aa31ba363cfa7SAME, 0 breachesThe record
Lab 06: Which model should review your code?github-hostedubuntu-latest0442d084f93aa888409bSAME, 0 breachesThe record
Lab 06: Which model should review your code?org-hostedorg-ci-heavy0442d084f93aa888409bSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestaa35e6db47c9ab93279dSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyaa35e6db47c9ab93279dSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?github-hostedubuntu-latestaa35e6db47c9ab93279dSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?org-hostedorg-ci-heavyaa35e6db47c9ab93279dSAME, 0 breachesThe record
Lab 05: Can an agent go rogue?laptopmacaa35e6db47c9ab93279dSAME, 0 breachesThe record

A record names a runner and a machine class by a word from a closed list, never a runner or host name. The record hash covers what ran, on which commit, and what it printed, not where or when. These are record checks; a repeated hash is not independent validation of the research conclusion.