Centralized xfail list for the Syntara test suites, maintained by @syntara-ci-stability.
Each .md file corresponds to a test suite (e.g. backend.md). Tests to mark
as xfail are listed as H1 headings with the full pytest node-id. The text below
each heading explains the reason:
# tests/e2e/workflows/test_wait_node.py::test_wait_node_zero_duration_fails
orchestration bug: _process_node_result swallows failed-status dictsSet SYNTARA_XFAIL_SOURCE to a base URL or local folder path. Both the backend
and frontend append their own filename (backend.md / playwright.md).
# URL (CI default)
export SYNTARA_XFAIL_SOURCE=https://raw.githubusercontent.com/syntara-orchestration/syntara-ci/refs/heads/main/
# Local folder
export SYNTARA_XFAIL_SOURCE=./path/to/xfail/The Makefile reads SYNTARA_XFAIL_SOURCE, appends backend.md, and passes the
result to pytest via --xfail-from-url:
# Use the CI xfail list
SYNTARA_XFAIL_SOURCE=https://raw.githubusercontent.com/…/ make test-e2e
# Use a local folder
SYNTARA_XFAIL_SOURCE=./xfail/ make test-e2e
# Disable xfail entirely (default when unset)
make test-e2eThe Playwright fixture reads SYNTARA_XFAIL_SOURCE, appends playwright.md,
and fetches the file once per worker. Matching tests are marked as expected
failures via test.fail(). Test identifiers use the format
file.spec.ts > Describe block > test name.
# Run with xfail list
SYNTARA_XFAIL_SOURCE=https://raw.githubusercontent.com/…/ npx playwright test
# Disable (default — no env var)
npx playwright test