APEX-Agents · Law
World434_DPM_02
APEX-Agents task World434_DPM_02 in AI Agents for Real Estate Lease Review. Compare dual-harness agent runs across models, scores, and public traces.
Task prompt
What the agent was asked to do
As of January 1, 2025, SecureBox Self Storage has not made any changes to the Emergency Exit Signage and Lighting. If the FSD inspection report is considered a fire safety direction, what is the maximum potential fine that Securebox Self Storage might receive under Hong Kong law? Reply to me here in a short message, giving the main parts of an email that I can review. Give a brief explanation of your reasoning in it. Assume that SecureBox Self Storage is a composite building.
Published trajectories
Agent runs on this task
Curated dual-harness runs (parsed + original sandbox). Best scored run per model.
| Model | Harness | Score | Result | Links |
|---|---|---|---|---|
| fireworks models Kimi K2 | dual | 7/7 | Pass | Run detailsPublic trace |
| Gemini 3.1 Pro | dual | 7/7 | Pass | Run detailsPublic trace |
| GPT-5.4 | dual | 6/7 | Fail | Run detailsPublic trace |
| GPT-5.4 mini | dual | 6/7 | Fail | Run detailsPublic trace |
| GPT-5.5 | dual | 6/7 | Fail | Run detailsPublic trace |
| GPT-5.4 nano | dual | 4/7 | Fail | Run detailsPublic trace |
Grading rubric
Rubric criteria
Runs are graded against these criteria. Open a run for model-specific verdicts.
States that the applicable law is the Fire Safety (Buildings) Ordinance (Cap. 572)
States that the penalty for failing to comply with a fire safety direction is a fine at Level 6
States that a fine at Level 6 is HK$100,000
States that a fine of HK$10,000 applies for each day or part of a day during which the failure continues after the expiry of the period specified in the direction
States that the compliance deadline for the Emergency Exit Signage and Lighting was 45 days from the inspection date of 15 September 2024
States the number of days of non-compliance after the deadline is 63 days
States that the maximum potential fine is HK$730,000