Earthly vs Source-to-Image

Earthly Lunar monitors software development activity and applies engineering policies across developer and AI workflows. It collects structured metadata from source repositories and CI/CD pipelines, then evaluates it against policies an organization defines. Lunar can be added to existing CI/CD pipelines without YAML changes or wholesale pipeline rewrites. Its guardrails can run while code is authored, on pull requests, and at deployment gates; enforcement can progress from draft and scoring to comments or blocking pull requests and deployments. Earthly lists more than 200 prebuilt guardrails for testing, security, compliance, and operational readiness. The integrations page lists 66 integrations, with Bash-script options for custom collectors and catalogers. Documentation includes a SQL API for engineering data and a CI agent that can run on self-hosted runners and is compatible with major CI vendors. Earthly says its services and databases are hosted on AWS in the United States, data is encrypted at rest and in transit, and it has SOC 2 Type II compliance. Earthfiles, an open-source portable build syntax, remains available. Earthly says the company was founded in 2020 and Lunar launched in 2026.

Source-to-Image is a free, Apache-2.0 toolkit for building reproducible container images from source code. It places source in a builder image, lets that container prepare the application, and outputs a ready-to-run image. A documented workflow can build from a Git repository without a Dockerfile. Builder images may provide assemble and run scripts, plus optional scripts for saving artifacts and explaining usage. Compatible builders can reuse artifacts from an earlier output when incremental settings and scripts are supplied. A .s2iignore file filters source files and directories, while .s2i/environment defines build variables that also appear in the output image. S2I can use an OnBuild strategy with images containing Dockerfile ONBUILD instructions. Examples cover Docker, OpenShift, UBI, Ruby, and WildFly. It runs on Linux, macOS, and Windows, with Docker 1.6+ and Go 1.7.1+ listed as dependencies.