| # DisasterBench | |
| DisasterBench is a benchmark for evaluating LLM structured planning over typed multi-agent disaster-response workflows. | |
| ## Benchmark Statistics | |
| - 233 planning tasks | |
| - 26 task-interface agents | |
| - Typed compatibility graph | |
| - Structured executable workflow plans | |
| ## Dataset Structure | |
| Each example contains: | |
| - `task_id` | |
| - `task_desc` | |
| - `structured_plan` | |
| ## Links | |
| - Code: https://github.com/TamuChen18/DisasterBench_Open |