CODESTRUCT reframes codebases as AST action spaces to fix brittle LLM code agents
burkov · x · 2026-10-16
LLM code agents today treat repos as unstructured text, applying edits via string matching that often fails on formatting drift, ambiguity, or repeated text — causing costly trial-and-error and missing patches. CODESTRUCT reframes the codebase as a structured action space where agents operate on named AST entities (functions, classes) instead of text spans. Evaluated on SWE-Bench Verified (500 real Python bug-fix tasks) and CodeAs… benchmarks to test whether syntax-aware operations improve reliability and efficiency.
More from coding & agent
- A vector DB is not LLM memory: long thread maps write/read paths, failure modes and evals — techNmak · 2026-10-16
- 10 GitHub repos to see what your AI agent really did, from Langfuse to claude-code-log — victor_explore · 2026-10-16
- Mosaic: multi-session agentic RL with hierarchical advantage estimation — LARK-Lab · 2026-10-16
- Recursive Environment Scaling tanks six models' success rates by 57–77 points — Lingxiao Jing · 2026-10-16
- Harvard's MPT: decentralized agent swarms learn self-orchestration via bidding — Harvard · 2026-10-16
- How LiteLLM routes one OpenAI-shaped API across every model provider — Mahmoud_Zalt · 2026-10-16