Benchmark Radar™
Star —

UnderSpecBench

Explore score history ↗

UnderSpecBench evaluates coding agents on DevOps tasks under varying instruction underspecification, measuring action-boundary violations such as wrong-target or over-scope actions.

Released
2026-07-02
Institutions
Hong Kong University of Science and Technology · Tongji University · Guangzhou HKUST Fok Ying Tung Research Institute
Evaluates
Code & Software, General AI, cs.SE
Openness
unknown
Importer
Claire Radar
Review status
unreviewed

UnderSpecBench paper

No reported scores are on record for this benchmark yet.

Which sources cite UnderSpecBench?

Related Code & Software, General AI, cs.SE benchmarks