agent/codex
1
0
Fork 0
mirror of https://github.com/pchuan98/codex.git synced 2026-07-01 00:31:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

2- Use string service tiers in session protocol (#20971)

Browse Source 
## Summary
- break service tier session/op/app-server protocol fields from the
closed enum to string tier ids
- send the service tier string directly through model requests, prewarm,
compaction, memories, and TUI/app-server turn starts
- regenerate app-server protocol JSON/TypeScript schemas, removing the
standalone ServiceTier TS enum

## Verification
- just fmt
- cargo check -p codex-core -p codex-app-server -p codex-tui
- just write-app-server-schema

---------

Co-authored-by: Codex <noreply@openai.com>
This commit is contained in:
Ahmed Ibrahim
2026-05-06 18:00:21 +03:00
committed by GitHub
Unverified
parent ebd9ec05b4
commit be1d3cff93
65 changed files with 403 additions and 570 deletions
Download Patch File Download Diff File Expand all files Collapse all files
codex-rs/core/src/client.rs
+10 -15
View File
@@ -73,7 +73,6 @@ use codex_otel::current_span_w3c_trace_context;
use codex_protocol::SessionId;
use codex_protocol::ThreadId;
use codex_protocol::config_types::ReasoningSummary as ReasoningSummaryConfig;
use codex_protocol::config_types::ServiceTier;
use codex_protocol::config_types::Verbosity as VerbosityConfig;
use codex_protocol::models::ResponseItem;
use codex_protocol::openai_models::ModelInfo;
@@ -150,7 +149,7 @@ pub(crate) const WEBSOCKET_CONNECT_TIMEOUT: Duration =
pub(crate) struct CompactConversationRequestSettings {
pub(crate) effort: Option<ReasoningEffortConfig>,
pub(crate) summary: ReasoningSummaryConfig,
pub(crate) service_tier: Option<ServiceTier>,
pub(crate) service_tier: Option<String>,
}
/// Session-scoped state shared by all [`ModelClient`] clones.
@@ -683,7 +682,7 @@ impl ModelClient {
model_info: &ModelInfo,
effort: Option<ReasoningEffortConfig>,
summary: ReasoningSummaryConfig,
service_tier: Option<ServiceTier>,
service_tier: Option<String>,
) -> Result<ResponsesApiRequest> {
let instructions = &prompt.base_instructions.text;
let input = prompt.get_formatted_input();
@@ -722,11 +721,7 @@ impl ModelClient {
store: provider.is_azure_responses_endpoint(),
stream: true,
include,
service_tier: match service_tier {
Some(ServiceTier::Fast) => Some("priority".to_string()),
Some(service_tier) => Some(service_tier.to_string()),
None => None,
},
service_tier,
prompt_cache_key,
text,
client_metadata: Some(HashMap::from([(
@@ -1168,7 +1163,7 @@ impl ModelClientSession {
session_telemetry: &SessionTelemetry,
effort: Option<ReasoningEffortConfig>,
summary: ReasoningSummaryConfig,
service_tier: Option<ServiceTier>,
service_tier: Option<String>,
turn_metadata_header: Option<&str>,
inference_trace: &InferenceTraceContext,
) -> Result<ResponseStream> {
@@ -1215,7 +1210,7 @@ impl ModelClientSession {
model_info,
effort,
summary,
service_tier,
service_tier.clone(),
)?;
let inference_trace_attempt = inference_trace.start_attempt();
inference_trace_attempt.record_started(&request);
@@ -1293,7 +1288,7 @@ impl ModelClientSession {
session_telemetry: &SessionTelemetry,
effort: Option<ReasoningEffortConfig>,
summary: ReasoningSummaryConfig,
service_tier: Option<ServiceTier>,
service_tier: Option<String>,
turn_metadata_header: Option<&str>,
warmup: bool,
request_trace: Option<W3cTraceContext>,
@@ -1321,7 +1316,7 @@ impl ModelClientSession {
model_info,
effort,
summary,
service_tier,
service_tier.clone(),
)?;
let mut ws_payload = ResponseCreateWsRequest {
client_metadata: response_create_client_metadata(
@@ -1453,7 +1448,7 @@ impl ModelClientSession {
session_telemetry: &SessionTelemetry,
effort: Option<ReasoningEffortConfig>,
summary: ReasoningSummaryConfig,
service_tier: Option<ServiceTier>,
service_tier: Option<String>,
turn_metadata_header: Option<&str>,
) -> Result<()> {
if !self.client.responses_websocket_enabled() {
@@ -1514,7 +1509,7 @@ impl ModelClientSession {
session_telemetry: &SessionTelemetry,
effort: Option<ReasoningEffortConfig>,
summary: ReasoningSummaryConfig,
service_tier: Option<ServiceTier>,
service_tier: Option<String>,
turn_metadata_header: Option<&str>,
inference_trace: &InferenceTraceContext,
) -> Result<ResponseStream> {
@@ -1530,7 +1525,7 @@ impl ModelClientSession {
session_telemetry,
effort,
summary,
service_tier,
service_tier.clone(),
turn_metadata_header,
/*warmup*/ false,
request_trace,
codex-rs/core/src/codex_thread.rs
+2 -3
View File
@@ -11,7 +11,6 @@ use codex_protocol::config_types::ApprovalsReviewer;
use codex_protocol::config_types::CollaborationMode;
use codex_protocol::config_types::Personality;
use codex_protocol::config_types::ReasoningSummary;
use codex_protocol::config_types::ServiceTier;
use codex_protocol::config_types::WindowsSandboxLevel;
use codex_protocol::error::CodexErr;
use codex_protocol::error::Result as CodexResult;
@@ -53,7 +52,7 @@ use codex_rollout::state_db::StateDbHandle;
pub struct ThreadConfigSnapshot {
pub model: String,
pub model_provider_id: String,
pub service_tier: Option<ServiceTier>,
pub service_tier: Option<String>,
pub approval_policy: AskForApproval,
pub approvals_reviewer: ApprovalsReviewer,
pub permission_profile: PermissionProfile,
@@ -91,7 +90,7 @@ pub struct CodexThreadTurnContextOverrides {
pub model: Option<String>,
pub effort: Option<Option<ReasoningEffort>>,
pub summary: Option<ReasoningSummary>,
pub service_tier: Option<Option<ServiceTier>>,
pub service_tier: Option<Option<String>>,
pub collaboration_mode: Option<CollaborationMode>,
pub personality: Option<Personality>,
}
codex-rs/core/src/compact.rs
+1 -1
View File
@@ -524,7 +524,7 @@ async fn drain_to_completed(
&turn_context.session_telemetry,
turn_context.reasoning_effort,
turn_context.reasoning_summary,
turn_context.config.service_tier,
turn_context.config.service_tier.clone(),
turn_metadata_header,
// Rollout tracing currently models remote compaction only; local compaction streams
// are left untraced until the reducer has a first-class local compaction lifecycle.
codex-rs/core/src/compact_remote.rs
+1 -1
View File
@@ -174,7 +174,7 @@ async fn run_remote_compact_task_inner_impl(
CompactConversationRequestSettings {
effort: turn_context.reasoning_effort,
summary: turn_context.reasoning_summary,
service_tier: turn_context.config.service_tier,
service_tier: turn_context.config.service_tier.clone(),
},
&turn_context.session_telemetry,
&compaction_trace,
codex-rs/core/src/compact_remote_v2.rs
+1 -1
View File
@@ -252,7 +252,7 @@ async fn run_remote_compaction_request_v2(
&turn_context.session_telemetry,
turn_context.reasoning_effort,
turn_context.reasoning_summary,
turn_context.config.service_tier,
turn_context.config.service_tier.clone(),
turn_metadata_header,
&InferenceTraceContext::disabled(),
)
codex-rs/core/src/config/config_tests.rs
+23
View File
@@ -62,6 +62,7 @@ use codex_model_provider_info::LMSTUDIO_OSS_PROVIDER_ID;
use codex_model_provider_info::OLLAMA_OSS_PROVIDER_ID;
use codex_model_provider_info::WireApi;
use codex_models_manager::bundled_models_response;
use codex_protocol::config_types::ServiceTier;
use codex_protocol::models::ActivePermissionProfile;
use codex_protocol::models::ActivePermissionProfileModification;
use codex_protocol::models::ManagedFileSystemPermissions;
@@ -7132,6 +7133,28 @@ async fn explicit_null_service_tier_override_sets_fast_default_opt_out() -> std:
Ok(())
}
#[tokio::test]
async fn legacy_fast_service_tier_override_uses_priority_request_value() -> std::io::Result<()> {
let fixture = create_test_fixture()?;
let config = Config::load_from_base_config_with_overrides(
fixture.cfg.clone(),
ConfigOverrides {
cwd: Some(fixture.cwd_path()),
service_tier: Some(Some("fast".to_string())),
..Default::default()
},
fixture.codex_home(),
)
.await?;
assert_eq!(
config.service_tier,
Some(ServiceTier::Fast.request_value().to_string())
);
Ok(())
}
#[tokio::test]
async fn fast_default_opt_out_notice_config_is_respected() -> std::io::Result<()> {
let fixture = create_test_fixture()?;
codex-rs/core/src/config/mod.rs
+15 -11
View File
@@ -402,8 +402,8 @@ pub struct Config {
/// Optional override of model selection.
pub model: Option<String>,
/// Effective service tier preference for new turns (`fast` or `flex`).
pub service_tier: Option<ServiceTier>,
/// Effective service tier request id preference for new turns.
pub service_tier: Option<String>,
/// Model used specifically for review sessions.
pub review_model: Option<String>,
@@ -1867,7 +1867,7 @@ pub struct ConfigOverrides {
pub permission_profile: Option<PermissionProfile>,
pub default_permissions: Option<String>,
pub model_provider: Option<String>,
pub service_tier: Option<Option<ServiceTier>>,
pub service_tier: Option<Option<String>>,
pub config_profile: Option<String>,
pub codex_self_exe: Option<PathBuf>,
pub codex_linux_sandbox_exe: Option<PathBuf>,
@@ -2735,16 +2735,20 @@ impl Config {
notices.fast_default_opt_out = Some(true);
None
}
None => config_profile.service_tier.or(cfg.service_tier),
None => config_profile
.service_tier
.or(cfg.service_tier)
.map(|service_tier| service_tier.request_value().to_string()),
};
let service_tier = match service_tier {
Some(ServiceTier::Fast) if features.enabled(Feature::FastMode) => {
Some(ServiceTier::Fast)
let service_tier = service_tier.and_then(|service_tier| {
match ServiceTier::from_request_value(&service_tier) {
Some(ServiceTier::Fast) => features
.enabled(Feature::FastMode)
.then(|| ServiceTier::Fast.request_value().to_string()),
Some(ServiceTier::Flex) => Some(ServiceTier::Flex.request_value().to_string()),
None => Some(service_tier),
}
Some(ServiceTier::Fast) => None,
Some(ServiceTier::Flex) => Some(ServiceTier::Flex),
None => None,
};
});
let compact_prompt = compact_prompt.or(cfg.compact_prompt).and_then(|value| {
let trimmed = value.trim();
codex-rs/core/src/guardian/tests.rs
+1 -1
View File
@@ -1894,7 +1894,7 @@ async fn guardian_review_surfaces_responses_api_errors_in_rejection_reason() ->
#[tokio::test]
async fn guardian_parallel_reviews_fork_from_last_committed_trunk_history() -> anyhow::Result<()> {
const TEST_STACK_SIZE_BYTES: usize = 2 * 1024 * 1024;
const TEST_STACK_SIZE_BYTES: usize = 4 * 1024 * 1024;
let handle =
std::thread::Builder::new()
codex-rs/core/src/session/config_lock.rs
+5 -2
View File
@@ -109,7 +109,10 @@ fn save_session_resolved_fields(sc: &SessionConfiguration, lock_config: &mut Con
lock_config.model = Some(sc.collaboration_mode.model().to_string());
lock_config.model_reasoning_effort = sc.collaboration_mode.reasoning_effort();
lock_config.model_reasoning_summary = sc.model_reasoning_summary;
lock_config.service_tier = sc.service_tier;
lock_config.service_tier = sc
.service_tier
.as_deref()
.and_then(codex_protocol::config_types::ServiceTier::from_request_value);
lock_config.instructions = Some(sc.base_instructions.clone());
lock_config.developer_instructions = sc.developer_instructions.clone();
lock_config.compact_prompt = sc.compact_prompt.clone();
@@ -280,7 +283,7 @@ mod tests {
sc.base_instructions = "catalog instructions".to_string();
sc.developer_instructions = Some("catalog developer instructions".to_string());
sc.compact_prompt = Some("catalog compact prompt".to_string());
sc.service_tier = Some(codex_protocol::config_types::ServiceTier::Flex);
sc.service_tier = Some("flex".to_string());
let lockfile = sc.to_config_lockfile_toml().expect("lock should serialize");
let lock = &lockfile.config;
codex-rs/core/src/session/mod.rs
+4 -4
View File
@@ -588,7 +588,7 @@ impl Codex {
.auth_cached()
.and_then(|auth| auth.account_plan_type());
let service_tier = get_service_tier(
config.service_tier,
config.service_tier.clone(),
config.notices.fast_default_opt_out.unwrap_or(false),
account_plan_type,
config.features.enabled(Feature::FastMode),
@@ -787,18 +787,18 @@ impl Codex {
}
fn get_service_tier(
configured_service_tier: Option<ServiceTier>,
configured_service_tier: Option<String>,
fast_default_opt_out: bool,
account_plan_type: Option<AccountPlanType>,
fast_mode_enabled: bool,
) -> Option<ServiceTier> {
) -> Option<String> {
if configured_service_tier.is_some() || fast_default_opt_out || !fast_mode_enabled {
return configured_service_tier;
}
account_plan_type
.is_some_and(is_enterprise_default_service_tier_plan)
.then_some(ServiceTier::Fast)
.then_some(ServiceTier::Fast.request_value().to_string())
}
fn is_enterprise_default_service_tier_plan(plan_type: AccountPlanType) -> bool {
codex-rs/core/src/session/session.rs
+14 -6
View File
@@ -1,6 +1,7 @@
use super::*;
use crate::goals::GoalRuntimeState;
use codex_protocol::SessionId;
use codex_protocol::config_types::ServiceTier;
use codex_protocol::permissions::FileSystemPath;
use codex_protocol::permissions::FileSystemSpecialPath;
use codex_protocol::protocol::ThreadSource;
@@ -42,7 +43,7 @@ pub(crate) struct SessionConfiguration {
pub(super) collaboration_mode: CollaborationMode,
pub(super) model_reasoning_summary: Option<ReasoningSummaryConfig>,
pub(super) service_tier: Option<ServiceTier>,
pub(super) service_tier: Option<String>,
/// Developer instructions that supplement the base instructions.
pub(super) developer_instructions: Option<String>,
@@ -136,7 +137,7 @@ impl SessionConfiguration {
ThreadConfigSnapshot {
model: self.collaboration_mode.model().to_string(),
model_provider_id: self.original_config_do_not_use.model_provider_id.clone(),
service_tier: self.service_tier,
service_tier: self.service_tier.clone(),
approval_policy: self.approval_policy.value(),
approvals_reviewer: self.approvals_reviewer,
permission_profile: self.permission_profile(),
@@ -182,8 +183,15 @@ impl SessionConfiguration {
if let Some(summary) = updates.reasoning_summary {
next_configuration.model_reasoning_summary = Some(summary);
}
if let Some(service_tier) = updates.service_tier {
next_configuration.service_tier = service_tier;
if let Some(service_tier) = updates.service_tier.clone() {
// TODO(aibrahim): Remove once v2 clients no longer send the legacy
// "fast" service tier value.
next_configuration.service_tier = service_tier.map(|service_tier| {
ServiceTier::from_request_value(&service_tier)
.map_or(service_tier, |service_tier| {
service_tier.request_value().to_string()
})
});
}
if let Some(personality) = updates.personality {
next_configuration.personality = Some(personality);
@@ -309,7 +317,7 @@ pub(crate) struct SessionSettingsUpdate {
pub(crate) windows_sandbox_level: Option<WindowsSandboxLevel>,
pub(crate) collaboration_mode: Option<CollaborationMode>,
pub(crate) reasoning_summary: Option<ReasoningSummaryConfig>,
pub(crate) service_tier: Option<Option<ServiceTier>>,
pub(crate) service_tier: Option<Option<String>>,
pub(crate) final_output_json_schema: Option<Option<Value>>,
/// Turn-local environment override. `None` inherits the sticky thread
/// environments stored on `SessionConfiguration`; `Some([])` explicitly
@@ -905,7 +913,7 @@ impl Session {
thread_name: session_configuration.thread_name.clone(),
model: session_configuration.collaboration_mode.model().to_string(),
model_provider_id: config.model_provider_id.clone(),
service_tier: session_configuration.service_tier,
service_tier: session_configuration.service_tier.clone(),
approval_policy: session_configuration.approval_policy.value(),
approvals_reviewer: session_configuration.approvals_reviewer,
permission_profile: session_configuration.permission_profile(),
codex-rs/core/src/session/tests.rs
+22 -5
View File
@@ -2891,7 +2891,7 @@ fn get_service_tier_defaults_enterprise_accounts_to_fast() {
Some(AccountPlanType::Enterprise),
/*fast_mode_enabled*/ true,
),
Some(ServiceTier::Fast)
Some(ServiceTier::Fast.request_value().to_string())
);
assert_eq!(
get_service_tier(
@@ -2900,7 +2900,7 @@ fn get_service_tier_defaults_enterprise_accounts_to_fast() {
Some(AccountPlanType::EnterpriseCbpUsageBased),
/*fast_mode_enabled*/ true,
),
Some(ServiceTier::Fast)
Some(ServiceTier::Fast.request_value().to_string())
);
assert_eq!(
get_service_tier(
@@ -2909,7 +2909,7 @@ fn get_service_tier_defaults_enterprise_accounts_to_fast() {
Some(AccountPlanType::Business),
/*fast_mode_enabled*/ true,
),
Some(ServiceTier::Fast)
Some(ServiceTier::Fast.request_value().to_string())
);
assert_eq!(
get_service_tier(
@@ -2918,7 +2918,7 @@ fn get_service_tier_defaults_enterprise_accounts_to_fast() {
Some(AccountPlanType::Team),
/*fast_mode_enabled*/ true,
),
Some(ServiceTier::Fast)
Some(ServiceTier::Fast.request_value().to_string())
);
assert_eq!(
get_service_tier(
@@ -2927,7 +2927,7 @@ fn get_service_tier_defaults_enterprise_accounts_to_fast() {
Some(AccountPlanType::SelfServeBusinessUsageBased),
/*fast_mode_enabled*/ true,
),
Some(ServiceTier::Fast)
Some(ServiceTier::Fast.request_value().to_string())
);
}
@@ -2980,6 +2980,23 @@ async fn session_settings_null_service_tier_update_clears_service_tier() {
assert_eq!(updated.service_tier, None);
}
#[tokio::test]
async fn session_settings_legacy_fast_service_tier_update_uses_priority_request_value() {
let session_configuration = make_session_configuration_for_tests().await;
let updated = session_configuration
.apply(&SessionSettingsUpdate {
service_tier: Some(Some("fast".to_string())),
..Default::default()
})
.expect("legacy fast service tier update should apply");
assert_eq!(
updated.service_tier,
Some(ServiceTier::Fast.request_value().to_string())
);
}
pub(crate) async fn make_session_configuration_for_tests() -> SessionConfiguration {
let codex_home = tempfile::tempdir().expect("create temp dir");
let config = build_test_config(codex_home.path()).await;
codex-rs/core/src/session/turn.rs
+7 -2
View File
@@ -73,6 +73,7 @@ use codex_hooks::HookEventAfterAgent;
use codex_hooks::HookPayload;
use codex_hooks::HookResult;
use codex_protocol::config_types::ModeKind;
use codex_protocol::config_types::ServiceTier;
use codex_protocol::error::CodexErr;
use codex_protocol::error::Result as CodexResult;
use codex_protocol::items::PlanItem;
@@ -693,7 +694,11 @@ async fn track_turn_resolved_config_analytics(
permission_profile_cwd: turn_context.cwd.to_path_buf(),
reasoning_effort: turn_context.reasoning_effort,
reasoning_summary: Some(turn_context.reasoning_summary),
service_tier: turn_context.config.service_tier,
service_tier: turn_context
.config
.service_tier
.as_deref()
.and_then(ServiceTier::from_request_value),
approval_policy: turn_context.approval_policy.value(),
approvals_reviewer: turn_context.config.approvals_reviewer,
sandbox_network_access: turn_context.network_sandbox_policy().is_enabled(),
@@ -1858,7 +1863,7 @@ async fn try_run_sampling_request(
&turn_context.session_telemetry,
turn_context.reasoning_effort,
turn_context.reasoning_summary,
turn_context.config.service_tier,
turn_context.config.service_tier.clone(),
turn_metadata_header,
&inference_trace,
)
codex-rs/core/src/session/turn_context.rs
+1 -1
View File
@@ -415,7 +415,7 @@ impl Session {
per_turn_config.model_reasoning_effort =
session_configuration.collaboration_mode.reasoning_effort();
per_turn_config.model_reasoning_summary = session_configuration.model_reasoning_summary;
per_turn_config.service_tier = session_configuration.service_tier;
per_turn_config.service_tier = session_configuration.service_tier.clone();
per_turn_config.personality = session_configuration.personality;
per_turn_config.approvals_reviewer = session_configuration.approvals_reviewer;
per_turn_config.permissions.permission_profile =
codex-rs/core/src/session_startup_prewarm.rs
+1 -1
View File
@@ -232,7 +232,7 @@ async fn schedule_startup_prewarm_inner(
&startup_turn_context.session_telemetry,
startup_turn_context.reasoning_effort,
startup_turn_context.reasoning_summary,
startup_turn_context.config.service_tier,
startup_turn_context.config.service_tier.clone(),
startup_turn_metadata_header.as_deref(),
)
.await?;
codex-rs/core/tests/common/test_codex.rs
+3 -3
View File
@@ -654,7 +654,7 @@ impl TestCodex {
prompt,
AskForApproval::Never,
PermissionProfile::Disabled,
Some(service_tier),
Some(service_tier.map(|service_tier| service_tier.request_value().to_string())),
/*environments*/ None,
)
.await
@@ -716,7 +716,7 @@ impl TestCodex {
prompt: &str,
approval_policy: AskForApproval,
permission_profile: PermissionProfile,
service_tier: Option<Option<ServiceTier>>,
service_tier: Option<Option<String>>,
environments: Option<Vec<TurnEnvironmentSelection>>,
) -> Result<()> {
self.submit_turn_with_context(
@@ -734,7 +734,7 @@ impl TestCodex {
prompt: &str,
approval_policy: AskForApproval,
permission_profile: PermissionProfile,
service_tier: Option<Option<ServiceTier>>,
service_tier: Option<Option<String>>,
environments: Option<Vec<TurnEnvironmentSelection>>,
) -> Result<()> {
let (sandbox_policy, permission_profile) =
codex-rs/core/tests/suite/agent_websocket.rs
+1 -1
View File
@@ -313,7 +313,7 @@ async fn websocket_v2_first_turn_drops_fast_tier_after_startup_prewarm() -> Resu
.features
.enable(Feature::ResponsesWebsocketsV2)
.expect("test config should allow feature update");
config.service_tier = Some(ServiceTier::Fast);
config.service_tier = Some(ServiceTier::Fast.request_value().to_string());
});
let test = builder.build_with_websocket_server(&server).await?;
codex-rs/core/tests/suite/client_websockets.rs
+1 -1
View File
@@ -1948,7 +1948,7 @@ async fn stream_until_complete_with_request_metadata(
&harness.session_telemetry,
harness.effort,
harness.summary,
service_tier,
service_tier.map(|service_tier| service_tier.request_value().to_string()),
turn_metadata_header,
&codex_rollout_trace::InferenceTraceContext::disabled(),
)
codex-rs/core/tests/suite/compact_remote.rs
+1 -1
View File
@@ -445,7 +445,7 @@ async fn assert_remote_manual_compact_request_parity(
let mut builder = test_codex().with_auth(auth);
if let Some(service_tier) = configured_service_tier {
builder = builder.with_config(move |config| {
config.service_tier = Some(service_tier);
config.service_tier = Some(service_tier.request_value().to_string());
});
}
let harness = TestCodexHarness::with_builder(builder).await?;