Complement support
We'd like to have a workflow to run the complement test suite against grapevine. It doesn't have great coverage, but is better than nothing and has caught regressions in conduit and conduwuit. Ideally we want:
- A script to run complement locally
- Track the test baseline so that we can distinguish between regressions and already-failing tests
- Some way to flag skipped or flaky tests
- Check for changes from the baseline in CI
Most of this can probably be cherry-picked from conduwuit. Running tests in CI will be a little tricky because complement relies heavily on docker, and our CI runner is already dockerized. Layering docker containers like this is always a nightmare.
Edited by Olivia Lee