Problem
PR #1128 (fix: per-message trace context propagation for persistent sessions) has no tests at all in the diff. There is zero new test code in the patch.
Impact
- The PR title claims a fix but no test verifies the fix works
- Failing-first methodology is structurally impossible to verify from this diff
- A real bug could ship undetected
What was audited
MCE audit Tcas1128 (2026-07-19) caught 3 HIGH attacks via F-C:
- The diff adds no tests at all - there is zero new test code in the patch
- The failing-first test methodology is structurally impossible to verify from this diff alone
- P3 (multiple messages handled) is overclaimed - the diff modifies both the single-message path and the AsyncIterable branch, but the actual inject_trace_into... logic is unclear
Recommended fix
- The author should add tests that:
a. Verify trace context is propagated for single messages
b. Verify trace context is propagated for multi-message streams
c. Verify trace context is NOT propagated incorrectly (e.g., between sessions)
- The test should fail on the pre-patch code
Cross-references
Problem
PR #1128 (fix: per-message trace context propagation for persistent sessions) has no tests at all in the diff. There is zero new test code in the patch.
Impact
What was audited
MCE audit Tcas1128 (2026-07-19) caught 3 HIGH attacks via F-C:
Recommended fix
a. Verify trace context is propagated for single messages
b. Verify trace context is propagated for multi-message streams
c. Verify trace context is NOT propagated incorrectly (e.g., between sessions)
Cross-references