What it is
SWE-Gym pairs software issues with Python repositories, executable environments and tests. It provides tasks and trajectories for developing software engineering agents and verifiers.
A code change must work inside a repository. The environment connects an agent’s edits to executable feedback rather than judging a patch by its appearance.
How an agent works in it
Sees
Issue descriptions, repository files and execution output.
Does
Inspect files, edit code and run commands.
Feedback
Task-specific test outcomes provide verifiable feedback.
Example task (illustrative): Resolve an issue, inspect failing tests, and submit a working patch.
Setup
Use the project’s task environments and agent integration. Provision repository dependencies and keep evaluation tasks separate from training.
Use it with BoltzLabs
BoltzLabs does not host this project. To record an agent working on one of its tasks:
- Prepare one task from the project: its starter files, instructions, dependencies and check command.
- In Evals, choose New eval, upload the starter folder or enter its repository, then set the task and check command.
- Choose BoltzLabs models or your own endpoint. Each run gets its own machine.
- When the run finishes, open it to read the agent’s trace, file changes and check result.
Projects that need a desktop or phone, self-hosted websites, a game or a simulator need that runtime set up separately; an eval machine does not include it. Machine time uses account credits; hosted model calls are billed separately.