Bug reports and feature requests go to the GitHub issue tracker. Please include a minimal reproducible example. For a question about an admissibility verdict, include the DAG string, the exposure and outcome, the control set, the verdict you expected and a reference for it.
Usage questions are welcome as issues tagged question.
Code contributions: fork, branch and open a pull request. Keep the core
free of dependencies (base R only; heavier tools belong in optional
engines and in Suggests). Any change to the admissibility engine needs a
test under tests/testthat/ that encodes a verdict citable to the
graphical-identification literature. Run R CMD check before
submitting.