blobfishai/erpbench-100-suite
Finance
Professional
100 executable Oracle-Fusion-shaped ERP workflows with deterministic ERPScore grading.
Run this task
CLI:
inspect eval inspect_harbor/blobfishai_erpbench_100_suite --model openai/gpt-5Python:
from inspect_ai import eval
from inspect_harbor import blobfishai_erpbench_100_suite
eval(blobfishai_erpbench_100_suite(), model="openai/gpt-5")Dataset information
| Harbor registry | blobfishai/erpbench-100-suite |
| Inspect task | blobfishai_erpbench_100_suite |
| Latest digest | sha256:26648e276ee6e190c73c343bf93cf77823718c9c565f7d92a330b4ddbe5fdc56 |
| Samples | 100 |
| Source | https://github.com/blobfishai/erp-agent-simulation |
See Task Parameters for the parameter set shared across all Harbor tasks.