FineEnvs/data-agent-harbor-test

🎯 Data Agent — Harbor (test) A held-out benchmark for data-analysis agents: 250 tasks, deliberately balanced across difficulty and leaning toward the harder end so it actually separates good agents from great ones. Each task hands your agent a real dataset and a question; it explores, computes, and answers — and every answer is checked deterministically, with no LLM judge. Packaged in Harbor format, ready to run. Where it comes from Built from the jupyter-agent… See the full description on the dataset page: https://huggingface.co/datasets/FineEnvs/data-agent-harbor-test.

<iframe src="https://silytics.si/embed/item/9af2b7b4-3ea8-4257-b210-0452bf88a07d" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.