CD pipeline fails in `tests/test_release.py::test_run_functional_suite`

[luisfponce]

Link

• Run: https://github.com/luisfponce/web_api_knowledge/actions/runs/23204316577/job/67435245494

Summary

The CD workflow fails in the Python Tests job when pytest executes tests/test_release.py.
The failure is currently reported as a generic subprocess return-code assertion in test_run_functional_suite, which masks the specific functional test case that failed.

Current behavior

tests/test_release.py::test_run_functional_suite runs a nested pytest process against tests/functional and only checks returncode == 0. When one or more functional tests fail, the top-level test fails with a generic assertion:

E       AssertionError: assert 1 == 0
E        +  where 1 = CompletedProcess(..., returncode=1).returncode

This makes triage slower because the root failing test is not surfaced clearly from this wrapper test.

Expected behavior

When functional tests fail in CD:

1. The failing functional test name and error details should be clearly visible in logs.
2. The release gate should still fail, but with actionable diagnostics.

Evidence

From run logs:

tests/test_release.py::test_run_functional_suite FAILED
...
>       assert result.returncode == 0
E       AssertionError: assert 1 == 0
...
tests/test_release.py:13: AssertionError
...
collecting ... collected 47 items

Relevant files:

• webapi/tests/test_release.py
• .github/workflows/cd.yml

Likely root cause

At least one test inside webapi/tests/functional fails in CI/CD environment.
The wrapper test in webapi/tests/test_release.py obscures first-failure details by asserting only on process exit code.

Reproduction

1. Run release wrapper locally from repository root:
   • cd webapi && pytest tests/test_release.py -v
2. Run functional suite directly to expose failing case:
   • cd webapi && pytest tests/functional -v -x

Proposed fixes

Short term

• Improve test_run_functional_suite diagnostics by printing captured stdout and stderr when return code is non-zero.
• Optionally run with -x and --maxfail=1 to surface first failure quickly.

Medium term

• Consider running tests/functional directly in workflow instead of invoking nested pytest from a pytest test file.
• Keep test_release.py only if there is strong justification for wrapper behavior.

Acceptance criteria

• CD failure logs show the specific failing test file and assertion in functional suite.
• A contributor can identify root cause from one workflow run without rerunning locally.
• tests/test_release.py no longer fails with only a generic subprocess assertion message.

Implementation checklist

• Capture and display subprocess stdout and stderr in webapi/tests/test_release.py when failing.
• Validate diagnostics quality by forcing one functional test failure on a test branch.
• Confirm CD logs now expose actionable failure context.
• Merge and verify next CD run readability.

Notes

Potentially related files to inspect after root-failure is known:

• webapi/tests/functional/test_auth_routes.py
• webapi/tests/functional/test_prompts_routes.py
• webapi/tests/functional/test_users_routes.py
• webapi/tests/functional/test_cd.py