Sandbox

The sandbox executes a set of commands to validate Sweep's changes using linters and type-checkers after every edit. By default we use Trunk (opens in a new tab), an opinionated super-linter that installs all the common formatters and linters for your codebase.

To get started with the setup run the following installation script in the background, then follow the instructions below to set up your config:

curl https://raw.githubusercontent.com/sweepai/sweep/main/bin/install_sweep_sandbox.sh | sh

Configuration

Go to your local clone of your repo. Create your sweep.yaml if it doesn't already exist and copy the following template into it:

# Sweep AI turns bug fixes & feature requests into code changes (https://sweep.dev)
# For details on our config file, check out our docs at https://docs.sweep.dev
 
# If you use this be sure to frequently sync your default branch(main, master) to dev.
branch: 'main'
# By default Sweep will read the logs and outputs from your existing Github Actions. To disable this, set this to false.
gha_enabled: True
# This is the description of your project. It will be used by sweep when creating PRs. You can tell Sweep what's unique about your project, what frameworks you use, or anything else you want.
# Here's an example: sweepai/sweep is a python project. The main api endpoints are in sweepai/api.py. Write code that adheres to PEP8.
description: ''
 
# Default Values: https://github.com/sweepai/sweep/blob/main/sweep.yaml

Sweep uses our own Sandbox base image based this Dockerfile (opens in a new tab), and first installs everything in the install section. Then on every edit, it runs the commands in the check section for the edited file. When it errors (non-zero return code), the logs get fed back into Sweep to fix the error, guaranteeing that every commit generates validated code.

By default we use Trunk, an opinionated super-linter that installs all the common formatters and linters for your codebase. You can set up and configure Trunk for yourself by following https://docs.trunk.io/get-started (opens in a new tab). This is equivalent to having the following sandbox config in your sweep.yaml:

sandbox:
  install:
    - trunk init
  check:
    - trunk fmt {file_path}
    - trunk check --fix {file_path}

However, you can set up custom commands for your Sandbox. Here are some example sweep.yaml sandbox configurations:

sandbox:
  install:
    - pip install -r requirements.txt
  check:
    - pylint --errors-only {file_path}

sandbox:
  install:
    - apt install python3.11 -y
    - pip install poetry
    - poetry env use python3.11
    - poetry install
  check:
    - poetry run pylint --errors-only {file_path}

sandbox:
  install:
    - yarn install --ignore-engines
  check:
    - yarn run prettier --write {file_path}
    - yarn run eslint {file_path}
    - yarn run tsc

We recommend starting with the formatters, followed by the linters and type-checkers. This is because formatters give clear error messages for syntax errors.

Testing

To ensure that your sandbox is set up properly, you can test it locally.

If you ran the installation script at the top, sweep-sandbox should be fully installed by now. Ensure that sweep-sandbox is installed properly by running sweep-sandbox --help. Then execute the following in the base of your repo in the following format:

sweep-sandbox FILE_PATH

Replacing FILE_PATH with the file or directory you want sweep-sandbox to run on. For example in the Sweep repo, we run sweep-sandbox sweepai to run pylint on the entire sweepai directory and sweep-sandbox sweepai/api.py to check the single file api.py.