Problem
Research questions and friction points this paper is trying to address.
Parallel LLM-Agent Development
Semantic Coordination
Interface Changes
Passes Alone, Fails Together: Benchmarking Semantic Coordination in Parallel LLM-Agent Development
🤖 AI Summary
研究解决并行LLM代理开发中因接口或规则变更导致的合并失败问题，通过创建stale基准测试来评估，并提出使用消息描述变更以减少干扰。
📝 Abstract
Parallel coding agents can produce patches that work alone but fail when merged. This happens when one agent changes an interface or rule that another agent still relies on. We study these failures with stale, a benchmark for semantic coordination. Our evaluation runs the same tests on each patch alone and on their combination, counting only failures introduced by combining the patches. We use three tiers: synthetic tasks with controlled interface changes, pairs of merged pull requests, and constructed tasks that use real Django helpers. Among 834 runs on 417 mined Django pairs, only one showed interference after correcting the grading procedure. On constructed tasks using 12 Django helpers, interference occurred in 97% of runs. A message describing the completed concurrent change recovered 82% of runs. Reviewed pull requests may contain few unresolved parallel changes, even when agents fail on controlled tasks using real code. The constructed failure rates do not estimate how often these problems occur in practice.
Innovation
Methods, ideas, or system contributions that make the work stand out.
semantic coordination
parallel LLM-Agent development
benchmarking
stale
concurrent change
🔎 Similar Papers
No similar papers found.
