Change headers and struct of rate limits

[aibrahim-oai]

No description provided.

[chatgpt-codex-connector]

Codex Review: Here are some suggestions.

codex/codex-rs/tui/src/chatwidget/tests.rs

Lines 380 to 418 in 0b363ce

	#[test]
	fn rate_limit_warnings_emit_thresholds() {
	let mut state = RateLimitWarningState::default();
	let mut warnings: Vec<String> = Vec::new();
	warnings.extend(state.take_warnings(10.0, 55.0));
	warnings.extend(state.take_warnings(55.0, 10.0));
	warnings.extend(state.take_warnings(10.0, 80.0));
	warnings.extend(state.take_warnings(80.0, 10.0));
	warnings.extend(state.take_warnings(10.0, 95.0));
	warnings.extend(state.take_warnings(95.0, 10.0));
	assert_eq!(
	warnings.len(),
	6,
	"expected one warning per threshold per limit"
	);
	assert!(
	warnings
	.iter()
	.any(|w| w.contains("Hourly usage exceeded 50%")),
	"expected hourly 50% warning"
	);
	assert!(
	warnings
	.iter()
	.any(|w| w.contains("Weekly usage exceeded 50%")),
	"expected weekly 50% warning"
	);
	assert!(
	warnings
	.iter()
	.any(|w| w.contains("Hourly usage exceeded 90%")),
	"expected hourly 90% warning"
	);
	assert!(
	warnings
	.iter()
	.any(|w| w.contains("Weekly usage exceeded 90%")),

[P1] Adjust warning tests for new rate-limit messages

The warning text was rephrased to "Heads up, you've used over …" in RateLimitWarningState, but the unit test still asserts that the warnings contain substrings such as "Hourly usage exceeded 50%" and "Weekly usage exceeded 50%". With the new wording none of these substrings appear, so rate_limit_warnings_emit_thresholds will now fail even when the runtime behaviour is correct. The test expectations should be updated to match the new messages or to check the thresholds in a wording-agnostic way.

---

Reply with @codex fix comments to fix any unresolved comments.

About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you open a pull request for review, mark a draft as ready, or comment "@codex review". If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex fix this CI failure" or "@codex address that feedback".

[chatgpt-codex-connector]

[P1] Use correct HTTP header names when parsing secondary limits

The new rate‑limit parsing still looks for headers named x-codex-secondry-* in several places. Given that the struct and surrounding code consistently use the fully spelled term “secondary”, these typos will fail to match the actual x-codex-secondary-* headers emitted by the server, and parse_rate_limit_snapshot will always return None. Downstream consumers (warning messages, TUI status) will silently lose rate‑limit information.

Useful? React with 👍 / 👎.

[aibrahim-oai]

will fix on the api side first.

[chatgpt-codex-connector]

Summary

• Corrected rate limit snapshot parsing to look up the properly spelled x-codex-secondary-* headers and updated the mock SSE response to emit the same names so rate-limit snapshots are captured again. core/src/client.rsL487-L493 core/tests/suite/client.rsL761-L768
• Updated the TUI rate-limit warning test to assert the new “Heads up…” copy for every threshold, ensuring the warning logic is covered end-to-end. tui/src/chatwidget/tests.rsL381-L413

Testing

• ✅ just fmt
• ✅ just fix -p codex-core
• ✅ just fix -p codex-tui
• ✅ USER=root cargo test -p codex-core
• ✅ cargo test -p codex-tui

View task →