snorkel-ai/senior-swe-bench-v2026.06

Senior SWE-Bench: senior-engineer coding tasks drawn from real pull requests in production repositories — building features from PM-style instructions and investigating bugs from runtime evidence, judged for correctness and code taste (50-task public split).

← Back to Registry

Run this task

CLI:

inspect eval inspect_harbor/snorkel_ai_senior_swe_bench_v2026_06 --model openai/gpt-5

Python:

from inspect_ai import eval
from inspect_harbor import snorkel_ai_senior_swe_bench_v2026_06

eval(snorkel_ai_senior_swe_bench_v2026_06(), model="openai/gpt-5")

Dataset information

Harbor registry	snorkel-ai/senior-swe-bench-v2026.06
Inspect task	snorkel_ai_senior_swe_bench_v2026_06
Latest digest	sha256:fdfbce8953be56c009a773140adccc53e24f92d8682e1bf8b2f785506a51f15d
Samples	50
Source	https://github.com/snorkel-ai/senior-swe-bench-v2026.06

See Task Parameters for the parameter set shared across all Harbor tasks.