Large language model agents have made strong progress on software engineering, yet current systems suffer from a context coupling problem: the standard code editing interface conflates code inspection, modification planning, and edit execution within a single context window, forcing agents to interleave exploratory viewing with strictly formatted edit generation. Irrelevant context accumulates and edit reliability degrades. We propose SWE-Edit, which decomposes the editing interface into two specialized subagents: a Viewer that extracts task-relevant code on demand, and an Editor that executes modifications from high-level natural language plans—letting the main agent focus on reasoning while delegating context-intensive operations to clean context windows. We further show that effective edit-format selection can be trained into a small model rather than requiring frontier-scale capacity: GRPO training on Qwen3-8B improves edit success by 12.5 pp and brings an 8B open-source editor to parity with GPT-5-nano on downstream SWE-Bench resolve rate, with the learned adaptive find-replace/whole-file-rewrite policy contributing primarily on the cost axis (-11.8% agent cost). To enable rapid editor iteration, we release PR-Edit, a lightweight editor-level evaluation that rank-orders the editor models we test consistently with their downstream SWE-Bench resolve rate. We release our code at https://github.com/microsoft/SWE-Edit.

Production LLM deployments maintain heterogeneous model pools spanning order-of-magnitude cost differences, yet existing routers make binary strong-vs-weak decisions coupled to specific model identities, requiring retraining on every catalog change. We present HyDRA (Hybrid Dynamic Routing Architecture), which uses a lightweight ModernBERT encoder to predict capability requirements (reasoning, code generation, debugging, and tool use) for each query. It then selects the cheapest model meeting those requirements through shortfall matching. HyDRA decouples the learned predictor from the model catalog, enabling catalog changes without retraining. On SWE-Bench Verified, HyDRA is 0.3 QR points below always-strong (not statistically significant) at 54.1% cost savings. On a held-out multilingual evaluation set, the same checkpoint retains 80.6% of oracle-routing quality at 37.5% cost savings, tunable to 79.1% retention at 56.1% savings. A 14-day A/B flight with close to 1M users per arm found statistically significant improvements in latency, time-to-first-token, reliability, and user activity. Deployed across GitHub Copilot VS Code Chat, GitHub Copilot CLI, and GitHub Copilot extensions, HyDRA demonstrates cross-lingual routing consistency across 16 languages and 4 language groups.

Code translation benchmarks are essential for evaluating the accuracy and efficiency of LLM-based systems. Existing benchmarks mainly target individual functions, overlooking repository-level challenges like intermodule coherence and dependency management. Recent repository-level efforts exist, but suffer from poor maintainability and coarse evaluation granularity. We introduce Skeleton-Guided-Translation, a framework for benchmarking Java-to-C# translation at the repository level, featuring fine-grained quality evaluation. It follows a two-step process: first translating repository “skeletons”, then refining the entire repository guided by these skeletons. Based on this, we present TRANSREPO-BENCH , the first test-driven benchmark of high-quality Java repositories paired with C# skeletons, unit tests, and build configurations. Our adaptive unit tests support multiple and incremental translations without manual tuning, enhancing automation and scalability. We also propose fine-grained metrics that evaluate translation quality per test case, overcoming limitations of binary metrics in distinguishing build failures. Evaluations using TRANSREPO-BENCH reveal issues like broken cross-file references, showing that our structured approach reduces dependency errors and preserves interface consistency.

Code execution is a fundamental aspect of programming language semantics that reflects the exact behavior of the code. However, most pre-trained models for code intelligence ignore the execution trace and only rely on source code and syntactic structures. In this paper, we investigate how well pre-trained models can understand and perform code execution. We develop a mutation-based data augmentation technique to create a large-scale and realistic Python dataset and task for code execution, which challenges existing models such as Codex. We then present CodeExecutor, a Transformer model that leverages code execution pre-training and curriculum learning to enhance its semantic comprehension. We evaluate CodeExecutor on code execution and show its promising performance and limitations. We also demonstrate its potential benefits for code intelligence tasks such as zero-shot code-to-code search and text-to-code generation. Our analysis provides insights into the learning and generalization abilities of pre-trained models for code execution.