gcp.rs test encodes timestamp substitution as correct behavior — a known data-integrity bug has a passing test validating it

[ooloth]

Current state

clients/src/gcp.rs parse_entries silently substitutes Utc::now() when a log entry's timestamp field is absent or unparseable, making stale entries appear recent in the TUI. This is a known data-integrity bug. The test parse_entries_missing_timestamp_uses_now exists and passes — it asserts entries[0].timestamp >= before && entries[0].timestamp <= after, which validates the silent substitution as the correct contract. The test gives a false green: it passes when the implementation is broken, and would continue to pass even if the substitution were removed and replaced with an error return. A developer trying to fix the bug cannot use this test to drive the fix — it will pass either way.

Ideal state

• The test either asserts parse_entries returns Err when a timestamp is absent (if that is the correct contract) or is marked #[ignore] with a comment linking to the bug issue and explaining that the current behavior is a known defect
• The test does not encode the silent substitution as a permanent behavioral contract
• When the underlying bug is fixed, the test is updated to match the corrected behavior

Out of scope

• Fixing the parse_entries implementation (that is the subject of the separate bug issue tracking the timestamp substitution)

Starting points

• clients/src/gcp.rs lines 182–190 — parse_entries_missing_timestamp_uses_now test

QA plan

1. Open clients/src/gcp.rs — expect parse_entries_missing_timestamp_uses_now to either assert Err or be marked #[ignore] with an explanatory comment referencing the bug issue
2. Run cargo test -p clients — expect the suite to pass (either the assertion is now correct, or the test is explicitly ignored)

Done when

parse_entries_missing_timestamp_uses_now does not assert that silent Utc::now() substitution is the correct behavioral contract.