agent-compat is an open-source SDK for detecting agent environments, compiling manifests into native files, and validating generated output against specifications. It is designed for teams building with multiple AI agents and developer tools. Why it exists Agent environments do not all consume configuration the same way. Maintaining equivalent rules or instructions manually creates duplication and drift.
I built agent-compat because agent configuration is starting to look like a compilation problem: one source of intent, multiple target environments, and a need for validation.
The first version is intentionally focused. I want to learn which compatibility edge cases developers hit in real projects before expanding the adapter surface.
Try it, open an issue, or tell me what your current workflow looks like.
GitHub: https://github.com/JustineDevs/a...
Report
Maker
The companion CLI is useful, but the main project is an SDK/library. The goal is to make cross-agent compatibility embeddable in tools, CI pipelines, and developer workflows rather than limit it to a standalone command.
The companion CLI is useful, but the main project is an SDK/library. The goal is to make cross-agent compatibility embeddable in tools, CI pipelines, and developer workflows rather than limit it to a standalone command.
What should the first community adapter support?