Recognises a toxic combination across a job's roles
A job holding two roles that together break an SoD rule.
Can AI do this?
Not settled yet The benchmark tested this, but the models' answers do not settle yet whether they master it. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 88% | Not shown |
| gemini-3.5-flash-lite | Low cost | 21% | Not shown |
| gpt-6-luna | Low cost | 89% | Not shown |
| claude-haiku-5-5 | Low cost | 89% | Not shown |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: The job, the roles and the rule
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: JB-003 payments operator holds RL-003 (create) and RL-008 (approve) Acceptable: JB-003 holds RL-003 only; JB-004 holds RL-008
Also known as: Toxic combination across roles
Related knobits
Builds on
All 20 knobits of Design access control.
Source: Review checklist 03-access-control, row 3 (27002 5.3)