DeepYardDeepYard
R

RepoLaunch

Automated repository testing agent for cross-language SWE dataset construction

Open SourceFree

About

First LLM agent capable of autonomously resolving dependencies, compiling source code, and extracting test results across arbitrary programming languages and operating systems. Designed specifically for automated software engineering (SWE) dataset construction, eliminating manual setup traditionally required for repository evaluation. Enables researchers to build large-scale code testing benchmarks without language-specific tooling.

Details

Typecoding-agent
Deploymentself-hosted
Supported Models

Tags

autonomouscoding-agentevaluationopen-sourcemulti-agentcli