APEX-Agents · Law
World433_MMF_02
APEX-Agents task World433_MMF_02 in AI Agents for Healthcare and Senior Living Legal Risk. Compare dual-harness agent runs across models, scores, and public traces.
Task prompt
What the agent was asked to do
Review Dr. Anderson's expert report and analyze whether (1) her opinions as to cause of death are admissible and (2) whether you believe there is a greater than an 80% chance of surviving a summary judgment motion on a crumbling skull rule affirmative defense to the complaint if her report is admitted. The case will be heard in New York. Support your assertions with facts from the case files. Provide your answer in a new documents document that you create.
Published trajectories
Agent runs on this task
Curated dual-harness runs (parsed + original sandbox). Best scored run per model.
| Model | Harness | Score | Result | Links |
|---|---|---|---|---|
| GPT-5.4 nano | dual | 5/8 | Fail | Run detailsPublic trace |
| fireworks models Kimi K2 | dual | 3/8 | Fail | Run detailsPublic trace |
| GPT-5.5 | dual | 3/8 | Fail | Run detailsPublic trace |
| GPT-5.4 | dual | 2/8 | Fail | Run detailsPublic trace |
| GPT-5.4 mini | dual | 2/8 | Fail | Run detailsPublic trace |
| Gemini 3.1 Pro | dual | 0/8 | Fail | Run detailsPublic trace |
Grading rubric
Rubric criteria
Runs are graded against these criteria. Open a run for model-specific verdicts.
States that Dr Anderson claims that Mr. Buchanan's death was triggered by an unknown event rather than a missed insulin dose
States that Dr Anderson claims that Tom Buchanan’s demise is "generally" triggered by an unknown psychological stressor
States that Dr Anderson claims that Tom Buchanan’s demise is generally not triggered by a singular medication omission
States that Dr Anderson's opinion as to cause of death is insufficient as a matter of law to preclude summary judgment on the crumbling skull rule
States that Dr Anderson's speculative statements on undiagnosed stressors are not admissible evidence
States that affidavits devoid of evidentiary facts are insufficient to defeat a motion for summary relief
States that Dr Anderson does not specify the undiagnosed event that caused Tom Buchanan's death
States that there is not a greater than 80% chance of surviving summary judgment