No description available

Docker image registry for SWE-bench, created by Epoch AI.

Hard-level DevOps/SWE terminal task built using the Terminal Bench 2.0 framework. This scenario requires deep reasoning across Linux systems, dependency resolution, runtime debugging, and environment reconstruction inside a Dockerized setup. Designed to evaluate problem-solving ability beyond surface-level file manipulation.

End-to-end TeamCity framework to run AI agents on SWE-Bench Lite. Spin up isolated Docker images per task, extract patches, score with the official harness, and aggregate success rates. As an example, we'll look at Junie and Google Gemini CLI

Heavily compressed docker images for SWE Bench Verified

Create, customize, and manage SWE-Bench containers

One-command SWE-bench eval harness in Go. Native ARM64 containers with 6.3x test runner speedup on Apple Silicon and AWS Graviton. Pre-built images on Docker Hub.

No description available

No description available

No description available

Utilities for running SWE-Bench with pre-built Docker containers locally

SWE-bench style coding agent evaluation framework using Modal sandboxes and Docker

Automated validation system for SWE-bench data points using Docker-based evaluation

Utilities for running SWE-Bench with pre-built Docker containers and Modal