EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

StreamSQL-Repair-Bench: Does Execution Feedback Let Agents Fix Event-Time Streaming SQL?

Zhuoxi Wang, Jingwen Hou

Fundamental Scientific Reports in Multidisciplinary Areas · 25.06.2026 · 2 (01) · 299-306

DOI: 10.66238/fsrma97 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Fundamental Scientific Reports in Multidisciplinary Areas, online erschienen am 25.06.2026. Verfasst von Zhuoxi Wang und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Writing a streaming query that runs is easy; writing one that is correct under event-time semantics is not. Watermarks and late-data policies are subtle, and the kind of temporal reasoning they demand is something large language models (LLMs) handle poorly. The dominant fix in agentic coding is execution feedback: let the model run its code, see what comes out, and revise. We ask whether that loop actually rescues streaming-SQL errors. StreamSQL-Repair-Bench hands a model a streaming-SQL query with a single planted bug, an intent description, and a few input/output tests, and asks it to repair the query; ground truth comes from a small reference engine implementing Dataflow-model event-time semantics, so we grade by execution accuracy without running a cluster. We separate bugs into shallow (readable straight off the I/O) and deep (require simulating event time), and compare one-shot repair, blind retrying, and repair with execution feedback. Across five LLMs, execution feedback barely moves shallow bugs (approx.0.9 either way) but roughly doubles deep-bug accuracy (e.g. Claude Haiku 0.14->0.53, Gemini Flash 0.28->0.56; aggregate 0.24->0.49). A blind-retry control stays at or below one-shot, so the gain comes from the executed result, not from extra attempts. Yet feedback is not a cure: it repairs structural mistakes (wrong window type or size) but leaves watermark and late-data bugs unfixed, the same semantics that resist one-shot reasoning. The benchmark, reference engine, and harness need no streaming infrastructure.

Zitieren

APA

Wang, Z., & Hou, J. (2026). StreamSQL-Repair-Bench: Does Execution Feedback Let Agents Fix Event-Time Streaming SQL?. Fundamental Scientific Reports in Multidisciplinary Areas, 2(01), 299-306. https://doi.org/10.66238/fsrma97

Vancouver

Wang Z, Hou J. StreamSQL-Repair-Bench: Does Execution Feedback Let Agents Fix Event-Time Streaming SQL?. Fundamental Scientific Reports in Multidisciplinary Areas. 2026;2(01):299-306. doi:10.66238/fsrma97

BibTeX
@article{Wang2026,
  author = {Wang, Zhuoxi and Hou, Jingwen},
  title = {StreamSQL-Repair-Bench: Does Execution Feedback Let Agents Fix Event-Time Streaming SQL?},
  journal = {Fundamental Scientific Reports in Multidisciplinary Areas},
  year = {2026},
  volume = {2},
  number = {01},
  pages = {299--306},
  doi = {10.66238/fsrma97}
}
Kostenlos starten →