Adam Perry @ OpenAI b7add4df3d Stabilize retry timing and WebSocket tests (#47437) ...

## Why

Retry timing assertions can fail if the test is descheduled before retry telemetry is emitted, and parallel tests can leave shared tracing callsites disabled for the test subscriber.

## What changed

- Measure retry intervals from the failed request, filter HTTP observations to `/responses`, and track WebSocket stream requests as well as connections.
- Keep a tracing interest-cache guard alive and rebuild callsite interest when installing the telemetry capture.
- Reserve the connection-failure test's port with a bound socket until the recovery server starts listening.
- Wait for the WebSocket warmup request and assert `generate: false` before submitting user input in connection-limit, rate-limit, and overload tests.

GitOrigin-RevId: bba39ea2ac29ac8553ae6184139382f1e6cdc7ac

2026-09-23 03:19:13 +00:00

..

skills_extension

Reuse cloud skill catalogs across turns until invalidated (#47350)

2026-09-22 19:58:52 +00:00

snapshots

Allow environment selection updates during an active turn (#47435)

2026-09-23 03:12:33 +00:00

step_settings

Allow environment selection updates during an active turn (#47435)

2026-09-23 03:12:33 +00:00

abort_tasks.rs

Route rollout reads through the canonical JSON decoder (#42378)

2026-09-02 22:45:19 +00:00

additional_context.rs

Unify context snapshots and group requests into windows (#44932)

2026-09-11 22:55:38 +00:00

agent_eviction_tests.rs

Prevent V2 agent eviction from racing with queued messages (#47240)

2026-09-22 09:50:48 +00:00

agent_execution.rs

Prevent V2 agent eviction from racing with queued messages (#47240)

2026-09-22 09:50:48 +00:00

agent_websocket.rs

Expose session analytics state in Responses turn metadata (#44628)

2026-09-10 19:41:14 +00:00

agents_md_refresh.rs

Add a provider for thread-scoped instructions (#44701)

2026-09-11 01:57:54 +00:00

agents_md.rs

Allow thread instruction providers to share updates with subagents (#46577)

2026-09-19 02:15:20 +00:00

app_tool_exposure.rs

Add per-app tool exposure configuration (#46035)

2026-09-16 21:50:07 +00:00

apply_patch_cli.rs

Enable MXC selection through Windows sandbox configuration (#46271)

2026-09-17 18:20:28 +00:00

apply_patch_serialization.rs

Standardize shell execution on unified exec (#39757)

2026-08-20 17:46:05 +00:00

approvals.rs

Remove retired model entries while preserving migration prompts (#44250)

2026-09-09 17:56:33 +00:00

audio_truncation.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

auth_recovery_policy.rs

Enforce application network policy across app-server requests (#47407)

2026-09-23 00:47:07 +00:00

auto_review.rs

Gate reasoning effort updates on explicit model support (#46530)

2026-09-19 00:18:56 +00:00

canonical_plugin_connectors.rs

Honor canonical plugin disables for shared connectors (#45755)

2026-09-15 18:23:08 +00:00

catalog_permission_messages.rs

Source async user message descriptions from the model catalog (#41461)

2026-08-29 03:44:01 +00:00

cli_stream.rs

Sanitize credentials from Git remote metadata (#40713)

2026-08-25 21:34:05 +00:00

client_websockets.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

client.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

cloud_config.rs

Preserve target-native paths in command approvals (#42577)

2026-09-03 16:21:10 +00:00

code_mode_elicitation.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

code_mode_model_messages.rs

Support model-catalog overrides for Code Mode tool messages (#47118)

2026-09-21 22:49:17 +00:00

code_mode.rs

Preserve empty cell metadata under pending tool-call pressure (#47174)

2026-09-22 03:03:52 +00:00

codex_apps_protocol.rs

Share Apps tool catalogs without retaining unused snapshots (#45440)

2026-09-14 15:06:28 +00:00

codex_delegate.rs

Preserve MCP tool output as content items (#40737)

2026-08-25 23:42:50 +00:00

collaboration_instructions.rs

Restore collaboration mode when resuming threads (#45519)

2026-09-14 21:33:45 +00:00

command_lifecycle_tests.rs

Add a command-start callback for tool lifecycle extensions (#46539)

2026-09-19 00:46:06 +00:00

compact_remote_trimming.rs

Estimate history tokens from content instead of serialized envelopes (#45094)

2026-09-12 18:41:47 +00:00

compact_remote.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

compact_resume_fork.rs

Remove the deprecated thread/rollback API (#44915)

2026-09-11 21:32:04 +00:00

compact.rs

Resume model context from the latest compaction boundary (#47365)

2026-09-22 21:17:08 +00:00

context_annotations.rs

Support image references by file ID in inputs and tool outputs (#45794)

2026-09-15 21:19:08 +00:00

current_time_reminder.rs

Report clock failures again after recovery (#46006)

2026-09-16 18:54:45 +00:00

cyber_access_program.rs

Remove the unused legacy remote compaction implementation (#44273)

2026-09-09 18:40:53 +00:00

cyber_exec_policy.rs

Standardize shell execution on unified exec (#39772)

2026-08-20 18:29:35 +00:00

daybreak_access.rs

Attach verified access context to eligible plugin MCP calls (#41005)

2026-08-27 00:02:26 +00:00

deprecation_notice.rs

Standardize shell execution on unified exec (#39772)

2026-08-20 18:29:35 +00:00

direct_tool_metadata.rs

Bind direct tool-call metadata to invocation outputs (#45185)

2026-09-13 06:53:20 +00:00

exec_policy.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

exec.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

extension_sandbox.rs

Expand extension permission regression coverage (#41946)

2026-09-01 00:42:19 +00:00

external_auth.rs

Add AWS credential export commands for Amazon Bedrock (#44028)

2026-09-09 03:17:48 +00:00

fork_thread.rs

Skip stored title lookups for ephemeral forks (#47342)

2026-09-22 19:00:27 +00:00

gateway_auth.rs

Compose gateway OAuth with primary provider authentication (#46490)

2026-09-18 22:42:46 +00:00

git_enrichment.rs

Establish root turn identity for independent tasks and memory requests (#42900)

2026-09-05 01:04:12 +00:00

guardian_authorization.rs

Support image references by file ID in inputs and tool outputs (#45794)

2026-09-15 21:19:08 +00:00

guardian_cached_score_tests.rs

Publish Guardian cached scores and coverage atomically (#46245)

2026-09-17 16:06:11 +00:00

guardian_checkpoint_migration_tests.rs

Separate thread startup metadata from replay history (#46294)

2026-09-17 19:55:20 +00:00

guardian_context_budget.rs

Enable Guardian thread context by default (#47275)

2026-09-22 12:43:11 +00:00

guardian_history.rs

Resume model context from the latest compaction boundary (#47365)

2026-09-22 21:17:08 +00:00

guardian_mcp_elicitation.rs

Preserve originating context for yielded code-mode tool calls (#44866)

2026-09-11 17:01:17 +00:00

guardian_persistence_tests.rs

Flush completed Guardian reviews before delivering decisions (#46509)

2026-09-18 23:28:13 +00:00

guardian_retained_context.rs

Resume model context from the latest compaction boundary (#47365)

2026-09-22 21:17:08 +00:00

guardian_retry.rs

Improve Guardian retries and review failure reporting (#44482)

2026-09-10 09:50:59 +00:00

guardian_review_cancellation.rs

Tighten request handling in Guardian approval tests (#46041)

2026-09-16 22:25:29 +00:00

guardian_review.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

guardian_sender_messages_tests.rs

Enable Guardian thread context by default (#47275)

2026-09-22 12:43:11 +00:00

guardian_subagent_authorization.rs

Preserve delivered messages through post-tool hook failures (#47039)

2026-09-21 14:14:21 +00:00

hooks_executor.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

hooks_mcp.rs

Set working directories for environment MCP tests (#41683)

2026-08-30 14:04:06 +00:00

hooks.rs

Preserve ordered assistant context in Guardian reviews (#47030)

2026-09-21 13:30:55 +00:00

image_rollout.rs

Pass thread IDs to attachment uploads (#47121)

2026-09-21 22:58:49 +00:00

injected_models_cache.rs

Scope model catalog caches to the current provider and auth identity (#43906)

2026-09-08 20:37:30 +00:00

interrupt_hooks.rs

Add hooks for interrupted turns (#40511)

2026-08-25 01:27:18 +00:00

items.rs

Support image references by file ID in inputs and tool outputs (#45794)

2026-09-15 21:19:08 +00:00

json_result.rs

Remove redundant test coverage (#42069)

2026-09-01 15:04:11 +00:00

live_cli.rs

[codex] Use expect in integration tests (#28441)

2026-06-15 21:53:47 -07:00

managed_threads_tests.rs

Decouple tool restrictions from session identity with ToolPolicy (#46999)

2026-09-21 10:40:40 +00:00

mcp_auth_elicitation.rs

Preserve originating turn metadata for yielded skill calls (#47106)

2026-09-21 20:34:46 +00:00

mcp_auth_refresh.rs

Allow subagents to request MCP elicitation input (#46877)

2026-09-20 19:42:18 +00:00

mcp_ema_config.rs

Resolve enterprise-managed MCP registrations in the catalog (#45459)

2026-09-14 16:30:00 +00:00

mcp_extension_protocol.rs

Preserve metadata in native user verification requests (#46952)

2026-09-21 07:01:52 +00:00

mcp_oauth_refresh_tests.rs

Preserve MCP authentication challenges on tool calls (#42552)

2026-09-03 15:12:02 +00:00

mcp_optional_startup_grace.rs

Make the optional MCP startup grace configurable (#41199)

2026-08-27 20:31:40 +00:00

mcp_refresh_cleanup.rs

Negotiate MCP extensions per app-server session (#36910)

2026-08-04 15:23:43 +00:00

mcp_startup_refresh_http_proxy.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

mcp_subagent_elicitation.rs

Allow subagents to request MCP elicitation input (#46877)

2026-09-20 19:42:18 +00:00

mcp_tool_cache.rs

Allow hosted Apps MCP contributions to override the protocol mode (#46019)

2026-09-16 20:32:52 +00:00

mcp_tool_exposure.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

mcp_turn_metadata.rs

Track cumulative MCP attribution across requests and thread history (#47081)

2026-09-21 18:19:07 +00:00

mcp_user_verification.rs

Restrict MCP user verification and add workspace-scoped identity (#43524)

2026-09-07 17:26:10 +00:00

mod.rs

Enforce application network policy across app-server requests (#47407)

2026-09-23 00:47:07 +00:00

model_overrides.rs

[5 of 7] Replace OverrideTurnContext with ThreadSettings (#22508)

2026-05-18 21:03:51 -07:00

model_provider_requirements_tests.rs

Connect app-server workspace discovery to model request routing (#46281)

2026-09-17 19:12:42 +00:00

model_runtime_selectors.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

model_switching.rs

Gate reasoning effort updates on explicit model support (#46530)

2026-09-19 00:18:56 +00:00

model_visible_layout.rs

Refresh host-supplied cloud skills at turn startup (#47074)

2026-09-21 17:28:14 +00:00

models_cache_auth.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

models_cache_ttl.rs

Gate reasoning effort updates on explicit model support (#46530)

2026-09-19 00:18:56 +00:00

models_etag_responses.rs

Standardize shell execution on unified exec (#39757)

2026-08-20 17:46:05 +00:00

multi_agent_mode.rs

Source proactive multi-agent instructions from the model catalog (#41457)

2026-08-29 03:28:19 +00:00

multi_agent_restore_tests.rs

Overlap agent metadata reads when resuming a V2 root (#47369)

2026-09-22 21:42:11 +00:00

multi_agent_resume.rs

Overlap agent metadata reads when resuming a V2 root (#47369)

2026-09-22 21:42:11 +00:00

multi_agent_tool_descriptions.rs

Support catalog parameter schemas for Multi-Agent V2 tools (#46505)

2026-09-18 23:16:56 +00:00

multi_exec_server_sandbox.rs

Attach verified access context to eligible plugin MCP calls (#41005)

2026-08-27 00:02:26 +00:00

network_approval_completion.rs

Add an opt-in preference for the local MXC sandbox (#47375)

2026-09-22 21:57:52 +00:00

network_approval.rs

Fix stale environment config in the network approval test (#46560)

2026-09-19 01:35:38 +00:00

openai_file_mcp.rs

Retry transient OpenAI file blob upload failures (#47393)

2026-09-22 23:56:23 +00:00

otel.rs

Remove selected core test cases (#42066)

2026-09-01 14:43:27 +00:00

override_updates.rs

Pair thread environment settings (#26687)

2026-06-08 13:55:15 -07:00

pending_input_persistence.rs

Overlap subagent spawn persistence and clean up cancelled children (#47348)

2026-09-22 19:39:18 +00:00

pending_input.rs

Add conditional turn interruption that preserves pending input (#47340)

2026-09-22 18:54:30 +00:00

permissions_messages.rs

Centralize model-message resolution and rendering in codex-prompts (#46026)

2026-09-16 21:01:13 +00:00

personality.rs

Retire the personality feature flag and document deprecated settings (#45809)

2026-09-15 23:39:40 +00:00

plugins.rs

Honor the configured product SKU in remote plugin requests (#47116)

2026-09-21 22:30:43 +00:00

prompt_cache_key.rs

Preserve parent cache affinity for ephemeral forks (#44862)

2026-09-11 16:41:56 +00:00

prompt_caching.rs

Centralize model-message resolution and rendering in codex-prompts (#46026)

2026-09-16 21:01:13 +00:00

prompt_debug_tests.rs

Enable git attribution across Codex entry points (#34819)

2026-07-22 20:23:35 +00:00

quota_exceeded.rs

Fix retry classification for throttling and quota errors (#45602)

2026-09-15 04:20:38 +00:00

realtime_conversation.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

realtime_initial_items.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

realtime_misalignment.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

realtime_sideband_endpoint.rs

Enforce application network policy across app-server requests (#47407)

2026-09-23 00:47:07 +00:00

realtime_system_proxy.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

reasoning_effort_override.rs

Preserve request-level reasoning effort for memory and title workers (#46531)

2026-09-19 00:19:20 +00:00

remote_env_failure_tests.rs

Surface environment startup failure reasons to the model (#44277)

2026-09-09 18:50:32 +00:00

remote_env.rs

Allow environment selection updates during an active turn (#47435)

2026-09-23 03:12:33 +00:00

remote_models.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

request_compression.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

request_permissions_tool.rs

Make permission transforms aware of executor path context (#41909)

2026-08-31 20:50:44 +00:00

request_permissions.rs

Make permission transforms aware of executor path context (#41909)

2026-08-31 20:50:44 +00:00

request_plugin_install.rs

Move plugin recommendations into developer context (#47349)

2026-09-22 19:51:33 +00:00

request_user_input_async.rs

Support catalog parameter schemas for Multi-Agent V2 tools (#46505)

2026-09-18 23:16:56 +00:00

request_user_input.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

responses_api_proxy_headers.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

responses_lite.rs

Support image references by file ID in inputs and tool outputs (#45794)

2026-09-15 21:19:08 +00:00

responses_system_proxy.rs

Add opt-in model discovery for OpenAI API keys (#44392)

2026-09-10 03:18:28 +00:00

resume_warning.rs

Preserve root turn attribution in turn-start events (#44611)

2026-09-10 18:24:13 +00:00

resume.rs

Separate thread startup metadata from replay history (#46294)

2026-09-17 19:55:20 +00:00

retry_after.rs

Stabilize retry timing and WebSocket tests (#47437)

2026-09-23 03:19:13 +00:00

review.rs

Preserve configured Flex tiers without catalog or fast-mode support (#46230)

2026-09-17 14:47:42 +00:00

rmcp_client.rs

Track cumulative MCP attribution across requests and thread history (#47081)

2026-09-21 18:19:07 +00:00

rollout_budget.rs

Remove the deprecated thread/rollback API (#44915)

2026-09-11 21:32:04 +00:00

rollout_compression.rs

Use StartThreadOptions across thread fork APIs (#44043)

2026-09-09 04:07:40 +00:00

rollout_list_find.rs

Honor the configured SQLite home across state consumers (#34994)

2026-07-23 19:19:35 +00:00

safety_buffering.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

safety_check_downgrade.rs

Preserve invalid_prompt as a distinct error classification (#47353)

2026-09-22 20:07:50 +00:00

scenarios_agent_message_board.rs

Add regression coverage for message board unsubscribe persistence (#47281)

2026-09-22 13:02:44 +00:00

scenarios_guardian_extra_policy.rs

Add extra policy configuration for Guardian reviews (#47125)

2026-09-21 23:06:36 +00:00

scenarios_mxc.rs

Add an opt-in preference for the local MXC sandbox (#47375)

2026-09-22 21:57:52 +00:00

scenarios_preparation.rs

Separate thread preparation from execution persistence (#47199)

2026-09-22 05:53:29 +00:00

scenarios_read_only_mcp.rs

Add read-only policy support to MCP tool requests (#46042)

2026-09-16 22:25:54 +00:00

scenarios_shared_instructions.rs

Keep Guardian reviews on the applied instruction snapshot (#46580)

2026-09-19 02:31:40 +00:00

scenarios.rs

Allow environment selection updates during an active turn (#47435)

2026-09-23 03:12:33 +00:00

search_tool.rs

Remove retired model entries while preserving migration prompts (#44250)

2026-09-09 17:56:33 +00:00

settings_commits.rs

Prevent V2 agent eviction from racing with queued messages (#47240)

2026-09-22 09:50:48 +00:00

settings_constraints.rs

Validate step settings against proposed permissions (#40647)

2026-08-25 17:27:04 +00:00

shell_snapshot.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

skill_approval.rs

Standardize shell execution on unified exec (#39757)

2026-08-20 17:46:05 +00:00

skills_extension.rs

Enforce application network policy for AWS auth and telemetry (#47408)

2026-09-23 00:47:34 +00:00

skills.rs

Refresh host-supplied cloud skills at turn startup (#47074)

2026-09-21 17:28:14 +00:00

spawn_agent_description.rs

Gate reasoning effort updates on explicit model support (#46530)

2026-09-19 00:18:56 +00:00

spawn_settings_tests.rs

Test spawn settings reporting after child removal (#47258)

2026-09-22 11:10:29 +00:00

sqlite_state.rs

Persist thread creator identity in rollouts and SQLite (#47113)

2026-09-21 21:22:01 +00:00

startup_cancellation.rs

Release persistent writers when session startup is cancelled (#44183)

2026-09-09 15:54:07 +00:00

step_settings_compaction.rs

Always use streamed remote compaction for supported providers (#44255)

2026-09-09 18:04:04 +00:00

step_settings_snapshots.rs

Remove retired model entries while preserving migration prompts (#44250)

2026-09-09 17:56:33 +00:00

step_settings.rs

Support model-catalog overrides for Code Mode tool messages (#47118)

2026-09-21 22:49:17 +00:00

stream_error_allows_next_turn.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

stream_no_completed.rs

Support explicit provider model catalog URLs (#46561)

2026-09-19 01:36:53 +00:00

subagent_notifications.rs

Overlap subagent spawn persistence and clean up cancelled children (#47348)

2026-09-22 19:39:18 +00:00

subagent_service_tier.rs

Make subagents follow the root service tier (#41308)

2026-08-28 09:48:13 +00:00

token_budget.rs

Refresh host-supplied cloud skills at turn startup (#47074)

2026-09-21 17:28:14 +00:00

token_usage_rollout.rs

Route rollout reads through the canonical JSON decoder (#42378)

2026-09-02 22:45:19 +00:00

tool_harness.rs

Update tests for default-enabled update_plan (#41630)

2026-08-30 07:40:36 +00:00

tool_lifecycle.rs

Add a command-start callback for tool lifecycle extensions (#46539)

2026-09-19 00:46:06 +00:00

tool_parallelism.rs

Standardize shell execution on unified exec (#39757)

2026-08-20 17:46:05 +00:00

tools.rs

Bind direct tool-call metadata to invocation outputs (#45185)

2026-09-13 06:53:20 +00:00

truncation.rs

Set working directories for environment MCP tests (#41683)

2026-08-30 14:04:06 +00:00

turn_input_submission.rs

Add opt-in reasoning status for realtime V3 delegations (#47377)

2026-09-22 22:20:00 +00:00

turn_state.rs

Preserve Markdown formatting when copying TUI responses (#42847)

2026-09-04 20:38:08 +00:00

unified_exec_process_events.rs

Handle pending network reviews after process completion (#42746)

2026-09-04 12:08:02 +00:00

unified_exec_stdin_approval.rs

Discount an approval's own code-mode wrapper from Guardian score lag (#43527)

2026-09-07 17:35:23 +00:00

unified_exec_stdin_review_size.rs

Reject NUL bytes in reviewed terminal input (#41354)

2026-08-28 15:44:48 +00:00

unified_exec_zsh_fork_approvals.rs

Keep step settings and approval environments consistent (#46556)

2026-09-19 01:27:26 +00:00

unified_exec.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

unstable_features_warning.rs

Extract persisted history types into a dedicated crate (#37871)

2026-08-10 19:26:52 +00:00

user_notification.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

user_shell_cmd.rs

Support configurable timeouts for thread shell commands (#41384)

2026-08-28 18:07:40 +00:00

view_image.rs

Gate reasoning effort updates on explicit model support (#46530)

2026-09-19 00:18:56 +00:00

web_search_system_proxy_tests.rs

Resolve proxy routes for standalone web search redirects (#47204)

2026-09-22 06:24:49 +00:00

web_search.rs

Add an Amazon Bedrock Runtime provider (#38470)

2026-08-14 01:23:59 +00:00

websocket_fallback.rs

Unify turn input submission and routing (#38275)

2026-08-13 00:00:16 +00:00

window_headers.rs

Use StartThreadOptions across thread fork APIs (#44043)

2026-09-09 04:07:40 +00:00

windows_sandbox.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

workspace_roots.rs

Always use private desktops for legacy Windows sandboxes (#46554)

2026-09-19 01:21:32 +00:00

worktree_trust.rs

Validate linked worktrees before inheriting project trust (#39616)

2026-08-20 06:29:32 +00:00