CodeScout: An Effective Recipe for Reinforcement Learning of Code Search Agents

This repository contains the source code for the paper **[CodeScout: An Effective Recipe for Reinforcement Learning of Code Search Agents](https://arxiv.org/abs/2603.17829)**.

🏆 CodeScout achieves open-source SOTA code localization performance outperforming 8-18x larger base and post-trained LLMs and narrows the gap with frontier closed-source models.

---

✨ Overview

A prerequisite for coding agents to perform tasks on large repositories is code localization - the identification of relevant files, classes, and functions to work on. While repository-level code localization has been performed using embedding-based retrieval approaches such as vector search, recent work has focused on developing agents to localize relevant code either as a standalone precursor to or interleaved with performing actual work. Most prior methods on agentic code search equip the agent with complex, specialized tools, such as repository graphs derived from static analysis. In [our paper](https://arxiv.org/abs/2603.17829), we demonstrate that, with an effective reinforcement learning recipe, a coding agent equipped with *nothing more* than a standard Unix terminal can be trained to achieve strong results. Our experiments on three benchmarks (SWE-Bench Verified, Pro, and Lite) r