@homeflare/distilled-litellm 0.2.0 → 0.3.0

package/src/services/config_yaml.ts

@@ -1,4 +1,4 @@
-// AUTO-GENERATED by scripts/generate.ts from .generated-specs (pinned litellm-openapi @ 1.100.0 — see README). Do not edit.
+// AUTO-GENERATED by scripts/generate.ts from .generated-specs (pinned litellm-openapi @ 1.103.0 — see README). Do not edit.
 import * as S from "@distilled.cloud/core/schema";
 import * as API from "@distilled.cloud/core/api";
 import * as C from "@distilled.cloud/core/category";
@@ -85,14 +85,26 @@ export const GetConfigGeneralSettingsConfigFieldInfoGetRequest =
     identifier: "GetConfigGeneralSettingsConfigFieldInfoGetRequest",
   }) as any as S.Schema<GetConfigGeneralSettingsConfigFieldInfoGetRequest>;
 
+export type ConfigFieldInfoSource =
+  | "config"
+  | "db"
+  | "env"
+  | "default"
+  | "unset";
+export const ConfigFieldInfoSource = S.String;
+
 export interface ConfigFieldInfo {
+  editable?: boolean;
   field_name: string;
   field_value: unknown;
+  source?: ConfigFieldInfoSource;
 }
 export const ConfigFieldInfo = /*@__PURE__*/ S.suspend(() =>
   S.Struct({
+    editable: S.optional(S.Boolean),
     field_name: S.String,
     field_value: S.Unknown,
+    source: S.optional(ConfigFieldInfoSource),
   }),
 ).annotate({
   identifier: "ConfigFieldInfo",
@@ -158,7 +170,11 @@ export const ConfigListNestedFieldsList = /*@__PURE__*/ S.Array(
   FieldDetail,
 ) as any as S.Schema<ConfigListNestedFieldsList>;
 
+export type ConfigListSource = "config" | "db" | "env" | "default" | "unset";
+export const ConfigListSource = S.String;
+
 export interface ConfigList {
+  editable?: boolean;
   field_default_value: unknown;
   field_description: string;
   field_name: string;
@@ -168,10 +184,12 @@ export interface ConfigList {
   field_value: unknown;
   nested_fields?: ConfigListNestedFieldsList | null;
   premium_field?: boolean;
+  source?: ConfigListSource;
   stored_in_db: boolean | null;
 }
 export const ConfigList = /*@__PURE__*/ S.suspend(() =>
   S.Struct({
+    editable: S.optional(S.Boolean),
     field_default_value: S.Unknown,
     field_description: S.String,
     field_name: S.String,
@@ -181,6 +199,7 @@ export const ConfigList = /*@__PURE__*/ S.suspend(() =>
     field_value: S.Unknown,
     nested_fields: S.optional(S.NullOr(ConfigListNestedFieldsList)),
     premium_field: S.optional(S.Boolean),
+    source: S.optional(ConfigListSource),
     stored_in_db: S.NullOr(S.Boolean),
   }),
 ).annotate({ identifier: "ConfigList" }) as any as S.Schema<ConfigList>;
@@ -226,6 +245,8 @@ export type AlertType =
   | "budget_alerts"
   | "spend_reports"
   | "failed_tracking_spend"
+  | "user_spend_thresholds"
+  | "user_spend_anomalies"
   | "db_exceptions"
   | "daily_reports"
   | "cooldown_deployment"
@@ -265,6 +286,12 @@ export const ConfigGeneralSettingsAlertingArgsMap = /*@__PURE__*/ S.Record(
   S.Unknown,
 ) as any as S.Schema<ConfigGeneralSettingsAlertingArgsMap>;
 
+export type ConfigGeneralSettingsAllowedFileExtensionsList = Array<string>;
+export const ConfigGeneralSettingsAllowedFileExtensionsList =
+  /*@__PURE__*/ S.Array(
+    S.String,
+  ) as any as S.Schema<ConfigGeneralSettingsAllowedFileExtensionsList>;
+
 export type ConfigGeneralSettingsAllowedRoutesList = Array<unknown>;
 export const ConfigGeneralSettingsAllowedRoutesList = /*@__PURE__*/ S.Array(
   S.Unknown,
@@ -277,6 +304,31 @@ export const ConfigGeneralSettingsBackgroundHealthCheckModelGroupsList =
     S.String,
   ) as any as S.Schema<ConfigGeneralSettingsBackgroundHealthCheckModelGroupsList>;
 
+export type ConfigGeneralSettingsBlockedFileExtensionsList = Array<string>;
+export const ConfigGeneralSettingsBlockedFileExtensionsList =
+  /*@__PURE__*/ S.Array(
+    S.String,
+  ) as any as S.Schema<ConfigGeneralSettingsBlockedFileExtensionsList>;
+
+export type ConfigGeneralSettingsClaudeCodeGatewayManagedSettingsMap = {
+  [key: string]: unknown | undefined;
+};
+export const ConfigGeneralSettingsClaudeCodeGatewayManagedSettingsMap =
+  /*@__PURE__*/ S.Record(
+    S.String,
+    S.Unknown,
+  ) as any as S.Schema<ConfigGeneralSettingsClaudeCodeGatewayManagedSettingsMap>;
+
+/** enable AWS ElastiCache IAM authentication */
+export type CoordinationRedisParamsAwsIamAuth = boolean | string;
+export const CoordinationRedisParamsAwsIamAuth =
+  S.Unknown as any as S.Schema<CoordinationRedisParamsAwsIamAuth>;
+
+/** the ElastiCache cache is serverless rather than a self-designed cluster */
+export type CoordinationRedisParamsAwsIamServerless = boolean | string;
+export const CoordinationRedisParamsAwsIamServerless =
+  S.Unknown as any as S.Schema<CoordinationRedisParamsAwsIamServerless>;
+
 export type CoordinationRedisParamsSentinelNodesItemItem = string | number;
 export const CoordinationRedisParamsSentinelNodesItemItem =
   S.Unknown as any as S.Schema<CoordinationRedisParamsSentinelNodesItemItem>;
@@ -318,6 +370,16 @@ export const CoordinationRedisParamsStartupNodesList = /*@__PURE__*/ S.Array(
 
 /** Connection params for the proxy's coordination Redis (cross-pod tpm/rpm rate limits, spend tracking, pod lock manager, shared health checks), configured independently of the response-cache backend in `litellm_settings.cache_params`. */
 export interface CoordinationRedisParams {
+  /** enable AWS ElastiCache IAM authentication */
+  aws_iam_auth?: CoordinationRedisParamsAwsIamAuth | null;
+  /** AWS ElastiCache cache name */
+  aws_iam_cache_name?: string | null;
+  /** AWS region for ElastiCache IAM authentication */
+  aws_iam_region?: string | null;
+  /** the ElastiCache cache is serverless rather than a self-designed cluster */
+  aws_iam_serverless?: CoordinationRedisParamsAwsIamServerless | null;
+  /** AWS ElastiCache IAM user name */
+  aws_iam_user_name?: string | null;
   /** Redis hostname */
   host?: string | null;
   /** Redis password */
@@ -341,6 +403,13 @@ export interface CoordinationRedisParams {
 }
 export const CoordinationRedisParams = /*@__PURE__*/ S.suspend(() =>
   S.Struct({
+    aws_iam_auth: S.optional(S.NullOr(CoordinationRedisParamsAwsIamAuth)),
+    aws_iam_cache_name: S.optional(S.NullOr(S.String)),
+    aws_iam_region: S.optional(S.NullOr(S.String)),
+    aws_iam_serverless: S.optional(
+      S.NullOr(CoordinationRedisParamsAwsIamServerless),
+    ),
+    aws_iam_user_name: S.optional(S.NullOr(S.String)),
     host: S.optional(S.NullOr(S.String)),
     password: S.optional(S.NullOr(S.String)),
     port: S.optional(S.NullOr(S.Number)),
@@ -434,6 +503,37 @@ export type KeyManagementSystem =
   | "custom";
 export const KeyManagementSystem = S.String;
 
+export type ConfigGeneralSettingsMaxFailedLoginAttemptsPerSourceOverridesMap = {
+  [key: string]: number | undefined;
+};
+export const ConfigGeneralSettingsMaxFailedLoginAttemptsPerSourceOverridesMap =
+  /*@__PURE__*/ S.Record(
+    S.String,
+    S.Number,
+  ) as any as S.Schema<ConfigGeneralSettingsMaxFailedLoginAttemptsPerSourceOverridesMap>;
+
+/** One entry of `general_settings.mcp_allowed_clients`. */
+export interface MCPAllowedClient {
+  /** Human-readable name for this client application, shown in the dashboard and in gateway logs. */
+  alias: string;
+  /** Exact value of the JWT claim named in litellm_jwtauth.mcp_client_id_jwt_field, or of the mcp_client_id_header header, that identifies this client application. Matched case-sensitively. */
+  value: string;
+}
+export const MCPAllowedClient = /*@__PURE__*/ S.suspend(() =>
+  S.Struct({
+    alias: S.String,
+    value: S.String,
+  }),
+).annotate({
+  identifier: "MCPAllowedClient",
+}) as any as S.Schema<MCPAllowedClient>;
+
+export type ConfigGeneralSettingsMcpAllowedClientsList =
+  Array<MCPAllowedClient>;
+export const ConfigGeneralSettingsMcpAllowedClientsList = /*@__PURE__*/ S.Array(
+  MCPAllowedClient,
+) as any as S.Schema<ConfigGeneralSettingsMcpAllowedClientsList>;
+
 export type ConfigGeneralSettingsMcpInternalIpRangesList = Array<string>;
 export const ConfigGeneralSettingsMcpInternalIpRangesList =
   /*@__PURE__*/ S.Array(
@@ -451,6 +551,12 @@ export const ConfigGeneralSettingsMcpTrustedProxyRangesList =
     S.String,
   ) as any as S.Schema<ConfigGeneralSettingsMcpTrustedProxyRangesList>;
 
+export type ConfigGeneralSettingsMissingSessionId =
+  | "generate"
+  | "reject"
+  | "omit";
+export const ConfigGeneralSettingsMissingSessionId = S.String;
+
 /** Key-value pairs of default query parameters to be sent with every request to this endpoint. These can be overridden by client-provided query parameters. For example: {'key': 'default_value', 'api_version': '2023-01'} */
 export type PassThroughGenericEndpointDefaultQueryParamsMap = {
   [key: string]: unknown | undefined;
@@ -643,7 +749,8 @@ export type SupportedDBObjectType =
   | "prompts"
   | "model_cost_map"
   | "tools"
-  | "config_overrides";
+  | "config_overrides"
+  | "websearch_interception_settings";
 export const SupportedDBObjectType = S.String;
 
 export type ConfigGeneralSettingsSupportedDbObjectsList = Array<
@@ -654,6 +761,12 @@ export const ConfigGeneralSettingsSupportedDbObjectsList =
     SupportedDBObjectType,
   ) as any as S.Schema<ConfigGeneralSettingsSupportedDbObjectsList>;
 
+export type ConfigGeneralSettingsTranscribeMediaBucketsList = Array<string>;
+export const ConfigGeneralSettingsTranscribeMediaBucketsList =
+  /*@__PURE__*/ S.Array(
+    S.String,
+  ) as any as S.Schema<ConfigGeneralSettingsTranscribeMediaBucketsList>;
+
 export type ConfigGeneralSettingsTrustedProxyRangesList = Array<string>;
 export const ConfigGeneralSettingsTrustedProxyRangesList =
   /*@__PURE__*/ S.Array(
@@ -700,11 +813,13 @@ export const ConfigGeneralSettingsUserUrlAllowedHostsList =
 
 /** Documents all the fields supported by `general_settings` in config.yaml */
 export interface ConfigGeneralSettings {
+  /** maximum time a request waits for a worker slot */
+  admission_queue_timeout_seconds?: number;
   /** Mapping of alert type to webhook url. e.g. `alert_to_webhook_url: {'budget_alerts': 'https://nothooks.slack.com/services/T00000000/B00000000/XXXXXXXXXXXXXXXXXXXXXXXX'}` */
   alert_to_webhook_url?: ConfigGeneralSettingsAlertToWebhookUrlMap | null;
   /** List of alerting types. By default it is all alerts */
   alert_types?: ConfigGeneralSettingsAlertTypesList | null;
-  /** List of alerting integrations. Today, just slack - `alerting: ['slack']` */
+  /** List of alerting integrations - e.g. `alerting: ['slack', 'webhook', 'email']`. 'slack' posts Slack-format messages to any Slack-compatible webhook (Slack, Rocket.Chat, Mattermost); 'webhook' posts structured JSON budget alerts to WEBHOOK_URL */
   alerting?: ConfigGeneralSettingsAlertingList | null;
   /** Controllable params for slack alerting - e.g. ttl in cache. */
   alerting_args?: ConfigGeneralSettingsAlertingArgsMap | null;
@@ -712,6 +827,10 @@ export interface ConfigGeneralSettings {
   alerting_threshold?: number | null;
   /** opt-in to RFC 8628 verification_uri_complete for the CLI SSO device flow, pre-filling the user_code in the browser. Off by default; intended for same-host clients where the device that starts the flow and the browser run on the same machine */
   allow_cli_sso_verification_uri_complete?: boolean | null;
+  /** If True, lets keys address Responses API ids that this proxy did not issue (raw provider ids, or ids issued before response-id encryption was configured). Such an id carries no owner, so no ownership check can run on it; ids this proxy did issue keep full ownership enforcement. Off by default, in which case an unrecognized response id is rejected with 403 */
+  allow_unmanaged_response_ids?: boolean | null;
+  /** the only file extensions (e.g. ['.jsonl', '.pdf', '.txt']) accepted on /v1/files uploads, for any purpose, matched case-insensitively against the uploaded filename. Files with any other extension, or none, are rejected. An empty list rejects every upload. Unset means no allowlist is applied */
+  allowed_file_extensions?: ConfigGeneralSettingsAllowedFileExtensionsList | null;
   /** Proxy API Endpoints you want users to be able to access */
   allowed_routes?: ConfigGeneralSettingsAllowedRoutesList | null;
   /** If True, a user's personal max_budget is enforced on every request they make, including requests made with a team-scoped key. Defaults to False, where a team-scoped key is governed only by the team and team-member budgets and the key owner's personal max_budget does not apply (see GitHub issue #12905). */
@@ -720,8 +839,12 @@ export interface ConfigGeneralSettings {
   background_health_check_model_groups?: ConfigGeneralSettingsBackgroundHealthCheckModelGroupsList | null;
   /** run health checks in background */
   background_health_checks?: boolean | null;
+  /** file extensions (e.g. ['.exe', '.sh']) rejected on /v1/files uploads, for any purpose, matched case-insensitively against the uploaded filename. Deprecated in favour of allowed_file_extensions; still enforced, after the allowlist, when set */
+  blocked_file_extensions?: ConfigGeneralSettingsBlockedFileExtensionsList | null;
   /** cancel the in-flight upstream LLM request (non-streaming) when the client disconnects, freeing backend capacity (e.g. a vLLM GPU slot); the request is logged as a 499 failure */
   cancel_on_disconnect?: boolean | null;
+  /** Claude Code managed-settings.json served verbatim at the gateway's /claude_code_gateway/managed/settings endpoint. When unset the endpoint returns 404 (no managed policy) */
+  claude_code_gateway_managed_settings?: ConfigGeneralSettingsClaudeCodeGatewayManagedSettingsMap | null;
   /** proxy level default model for all chat completion calls */
   completion_model?: string | null;
   /** Global Control Plane: URL of the control plane whose admin UI manages this instance. Enables /v3/login and /v3/login/exchange on this instance so that UI can authenticate against it cross-origin, and restricts the SSO return_to origin to that URL. No state is shared with the control plane */
@@ -742,7 +865,9 @@ export interface ConfigGeneralSettings {
   database_disable_prepared_statements?: boolean | null;
   /** Escape hatch: extra key/value pairs appended verbatim to the Prisma DATABASE_URL / DIRECT_URL query string (e.g. `sslmode`, `pgbouncer`, `statement_cache_size`). Keys here override any default LiteLLM sets. */
   database_extra_connection_params?: ConfigGeneralSettingsDatabaseExtraConnectionParamsMap | null;
-  /** Prisma `socket_timeout` URL param (seconds). When set, an idle/slow connection that has not produced data within this window is closed. This is the main knob for capping idle DB connections from LiteLLM. */
+  /** Prisma `max_idle_connection_lifetime` URL param (seconds). A pooled connection idle longer than this is closed and replaced instead of being handed to the next request. Defaults to 60 so connections are recycled before common infra idle timeouts (AWS NLB / RDS Proxy ~350s, many LBs 60-350s) silently drop them and requests fail with `Error { kind: Closed }`. A value pinned on the DATABASE_URL or set via `database_extra_connection_params` takes precedence. */
+  database_max_idle_connection_lifetime?: number | null;
+  /** Prisma `socket_timeout` URL param (seconds). When set, an in-flight operation that has not produced data within this window is aborted. For capping how long idle pooled connections are kept, see `database_max_idle_connection_lifetime`. */
   database_socket_timeout?: number | null;
   /** to use dynamodb instead of postgres db */
   database_type?: string | null;
@@ -750,12 +875,26 @@ export interface ConfigGeneralSettings {
   database_url?: string | null;
   /** By default, the user calling /team/new is automatically added to the new team as a team admin. If True, proxy admins are no longer auto-added; members explicitly listed in members_with_roles are unaffected. Default is False. */
   disable_auto_add_proxy_admin_to_teams?: boolean | null;
-  /** If True, disables the optimistic per-request budget reservation introduced in v1.84.0. WARNING: This weakens hard budget enforcement. Without the reservation, a burst of concurrent requests from a single key can each pass the read-time spend check before any of them is charged, allowing a configured budget to be exceeded under high concurrency. Budgets are still evaluated on every request at read time, so an already-exhausted budget is still rejected. Enable only if your deployment is experiencing phantom BudgetExceededError responses caused by leaked reservations (see GitHub issue #27639). A proxy-level WARNING is logged on every request while this flag is active as a reminder that hard enforcement is relaxed. */
+  /** If True, disables the optimistic per-request budget reservation introduced in v1.84.0. WARNING: This weakens hard budget enforcement. Without the reservation, a burst of concurrent requests from a single key can each pass the read-time spend check before any of them is charged, allowing a configured budget to be exceeded under high concurrency. Budgets are still evaluated on every request at read time, so an already-exhausted budget is still rejected. Enable only if your deployment is experiencing phantom BudgetExceededError responses caused by leaked reservations (see GitHub issue #27639). An INFO notice is logged once per worker at config load while this flag is active as a reminder that hard enforcement is relaxed. */
   disable_budget_reservation?: boolean | null;
+  /** If True, disables signing in to the Admin UI with the environment credentials: UI_USERNAME/UI_PASSWORD, or the master key when UI_PASSWORD is unset (that fallback means env-credential login is always live by default). Database users with passwords are unaffected. LOCKOUT RISK: create at least one proxy admin user with a password before enabling, or nobody can sign in to the UI. A locked-out admin can still administer the proxy over the API with the master key, and can unset this setting and restart the proxy to restore env-credential login. Default is False. */
+  disable_env_credential_login?: boolean | null;
+  /** If True and SSO is configured (MICROSOFT_CLIENT_ID, GOOGLE_CLIENT_ID, GENERIC_CLIENT_ID, or SAML_IDP_METADATA_URL/XML), disables username/password login on /login, /v2/login, and /v3/login so SSO is the only way to reach the Admin UI. An admin locked out of the UI can still administer the proxy over the API with the master key; unset this setting and restart the proxy to restore UI username/password login. Default is False. */
+  disable_password_login_when_sso_enabled?: boolean | null;
+  /** If True, disables ownership enforcement on Responses API ids. Keys may then retrieve, cancel, delete, and chain from any response id, including ids belonging to another user or team and ids this proxy never issued. WARNING: this removes tenant isolation on /v1/responses */
+  disable_responses_id_security?: boolean | null;
+  /** serve the Claude Code gateway protocol (https://code.claude.com/docs/en/claude-apps-gateway) under /claude_code_gateway: OAuth device-flow sign-in reusing proxy SSO, plus managed settings and OTLP telemetry ingestion. Off by default */
+  enable_claude_code_gateway?: boolean | null;
+  /** Serve the OpenAI pass-through WebSocket route, which relays frames to OpenAI under the proxy's own provider credential without reading them. Off by default. */
+  enable_openai_websocket_passthrough?: boolean | null;
   /** Public model hub for users to see what models they have access to, supported openai params, etc. */
   enable_public_model_hub?: boolean;
   /** If True, router fallbacks configured in router_settings are only attempted when the calling key (and its team and project) is allowed to call the fallback model; unauthorized fallback targets are skipped and the primary model's error is returned. Default is False. */
   enforce_fallback_model_access?: boolean | null;
+  /** How long a blocked source address, or source address and username, stays blocked. Every attempt from a blocked key, right or wrong, is refused with 429 before the password is checked; the block is not extended by refused attempts. Set under `general_settings` in config.yaml. Defaults to 300 */
+  failed_login_block_seconds?: number | null;
+  /** Fixed window in seconds over which failed Admin UI sign-in attempts are counted. The window starts at the first failure and is not extended by later ones. Set under `general_settings` in config.yaml. Defaults to 60 */
+  failed_login_window_seconds?: number | null;
   /** If True, forwards client headers (e.g. Authorization) to the LLM API. Required for Claude Code with Max subscription. */
   forward_client_headers_to_llm_api?: boolean | null;
   /** global max parallel requests to allow for a proxy instance. */
@@ -774,8 +913,18 @@ export interface ConfigGeneralSettings {
   master_key?: string | null;
   /** max batch input file size in MB for /v1/files uploads with purpose=batch, if a file is larger than this size it will be rejected before being forwarded to the provider */
   max_batch_file_size_mb?: number | null;
+  /** Failed Admin UI sign-in attempts allowed from one source address, across every username, within `failed_login_window_seconds`. One more blocks that address for `failed_login_block_seconds`. Half this value, rounded down but at least 1, is the allowance for one username from that address; one more blocks that address for that username only, and its further failures stop counting toward the address limit, so a script stuck on one account does not block everyone behind a shared address. The per-address limit is only enforced when `trusted_proxy_ranges` is set: to the proxies in front of LiteLLM, or to an empty list when clients connect directly. Left unset, the peer address may be a shared ingress and only the per-username half runs. IPv6 addresses are grouped by /64. Set under `general_settings` in config.yaml. Defaults to 10 */
+  max_failed_login_attempts_per_source?: number | null;
+  /** Per-address overrides of `max_failed_login_attempts_per_source`, keyed by IP address or CIDR range, e.g. {'1.2.3.4': 200, '5.6.0.0/24': 500}. The most specific matching range wins (between equivalent keys such as '1.2.3.4' and '1.2.3.4/32', an exemption wins, then the higher limit), and the per-username allowance for that address follows as half the override. A value of 0 exempts the address from both limits. Set under `general_settings` in config.yaml */
+  max_failed_login_attempts_per_source_overrides?: ConfigGeneralSettingsMaxFailedLoginAttemptsPerSourceOverridesMap | null;
+  /** max file size in MB for /v1/files uploads, for any purpose, if a file is larger than this size it will be rejected before being forwarded to the provider */
+  max_file_size_mb?: number | null;
+  /** maximum concurrent requests handled by each worker */
+  max_in_flight_requests_per_worker?: number | null;
   /** maximum parallel requests for each api key */
   max_parallel_requests?: number | null;
+  /** maximum requests waiting for a worker slot */
+  max_queued_requests_per_worker?: number | null;
   /** max request size in MB, if a request is larger than this size it will be rejected */
   max_request_size_mb?: number | null;
   /** max response size in MB, if a response is larger than this size it will be rejected */
@@ -794,6 +943,10 @@ export interface ConfigGeneralSettings {
   maximum_spend_logs_cleanup_run_budget?: string | null;
   /** Maximum retention period for spend logs (e.g., '7d' for 7 days). Logs older than this will be deleted. */
   maximum_spend_logs_retention_period?: string | null;
+  /** MCP client applications admitted by the gateway, each an {alias, value} pair where alias is the name shown in the dashboard and logs and value is the identity that must match exactly. When set, every MCP request must carry a client identity equal to one of the values: a JWT caller is identified by the claim named in litellm_jwtauth.mcp_client_id_jwt_field, any other caller by the header named in mcp_client_id_header. A request with no resolvable identity, or an unlisted one, is rejected with 403. Unset means every client is admitted. */
+  mcp_allowed_clients?: ConfigGeneralSettingsMcpAllowedClientsList | null;
+  /** Request header whose value names the calling MCP client application (for example 'x-mcp-client') for callers that did not authenticate with a JWT, used only while mcp_allowed_clients is set. The client picks this value itself, so it is a policy control rather than a security boundary; prefer litellm_jwtauth.mcp_client_id_jwt_field where callers use JWTs. */
+  mcp_client_id_header?: string | null;
   /** Custom CIDR ranges that define internal/private networks for MCP access control. When set, only these ranges are treated as internal. Defaults to RFC 1918 private ranges (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 127.0.0.0/8). */
   mcp_internal_ip_ranges?: ConfigGeneralSettingsMcpInternalIpRangesList | null;
   /** List of MCP server fields that must be filled in for a submission to pass standards checks (e.g. ['description', 'source_url', 'alias']). */
@@ -802,6 +955,11 @@ export interface ConfigGeneralSettings {
   mcp_trusted_proxy_ranges?: ConfigGeneralSettingsMcpTrustedProxyRangesList | null;
   /** Number of trusted reverse proxies/load balancers in front of the gateway that append to X-Forwarded-For. When set (and mcp_trusted_proxy_ranges validates the direct peer), the client IP for MCP access control is read this many entries from the right of the chain instead of the spoofable leftmost value, defeating append-style X-Forwarded-For forgery. */
   mcp_xff_num_trusted_hops?: number | null;
+  /** What to do with LLM API requests that carry no session id (x-litellm-session-id header, metadata.session_id, etc.). 'generate' stamps one id into litellm_session_id, litellm_trace_id and metadata.session_id so SpendLogs and logging callbacks agree; 'reject' returns 400; 'omit' leaves SpendLogs.session_id null, matching callbacks such as Langfuse that only record a client-established metadata.session_id. Unset keeps the legacy behavior where SpendLogs falls back to the trace id while callbacks get no session id. */
+  missing_session_id?:
+    | ConfigGeneralSettingsMissingSessionId
+    | (string & {})
+    | null;
   /** When true, `/models`, `/v1/models/{id}` and `/model/info` hide models whose backing deployments are all unhealthy, for every caller, without needing `healthy_only=true` per request. Requires `background_health_checks: true`, and keeps deployment health state cached without turning on `enable_health_check_routing`, so routing is unaffected. With no health state nothing is hidden. Hiding is presentation-only, a hidden model can still be called. */
   model_list_healthy_only?: boolean | null;
   /** [BETA] OpenTelemetry support - this might change, use with caution. */
@@ -810,6 +968,16 @@ export interface ConfigGeneralSettings {
   pass_through_endpoints?: ConfigGeneralSettingsPassThroughEndpointsList | null;
   /** Default upstream request timeout in seconds for native and custom pass-through endpoints that use pass_through_request. Defaults to 600 when unset. */
   pass_through_request_timeout?: number | null;
+  /** Minimum length required for a locally-managed user's password. Default is 12; a value below 8 is floored to 8 rather than weakening the requirement further. */
+  password_policy_min_length?: number | null;
+  /** If True (default), a locally-managed user's password must contain a lowercase letter. */
+  password_policy_require_lowercase?: boolean | null;
+  /** If True (default), a locally-managed user's password must contain a number. */
+  password_policy_require_numbers?: boolean | null;
+  /** If True (default), a locally-managed user's password must contain a special (non-alphanumeric) character. */
+  password_policy_require_special_characters?: boolean | null;
+  /** If True (default), a locally-managed user's password must contain an uppercase letter. */
+  password_policy_require_uppercase?: boolean | null;
   /** external services registered as embeddable UI plugins */
   plugins?: ConfigGeneralSettingsPluginsList | null;
   /** Allowlist of hosts a request may redirect a provider call's destination URL to. */
@@ -826,7 +994,9 @@ export interface ConfigGeneralSettings {
   store_prompts_in_spend_logs?: boolean | null;
   /** Fine-grained control over which object types to load from the database when store_model_in_db is True. Available types: 'models', 'mcp', 'guardrails', 'vector_stores', 'pass_through_endpoints', 'prompts', 'model_cost_map', 'tools', 'config_overrides'. If not set, all objects are loaded (default behavior). */
   supported_db_objects?: ConfigGeneralSettingsSupportedDbObjectsList | null;
-  /** CIDR ranges of trusted reverse proxies allowed to provide identity headers for header-based auth paths such as enable_oauth2_proxy_auth and custom_ui_sso_sign_in_handler. */
+  /** S3 bucket names that keys other than proxy admins may read media from and write transcripts to through the Amazon Transcribe pass-through. Unset means only proxy admins can start transcription jobs. */
+  transcribe_media_buckets?: ConfigGeneralSettingsTranscribeMediaBucketsList | null;
+  /** CIDR ranges of trusted reverse proxies allowed to provide identity headers for header-based auth paths such as enable_oauth2_proxy_auth and custom_ui_sso_sign_in_handler, and whose X-Forwarded-For is used to attribute Admin UI sign-in attempts to a source address. Set it to an empty list when clients connect directly, so the peer address is the source. Left unset, or containing an entry that is not an address or CIDR range, the per-source sign-in limit is off. */
   trusted_proxy_ranges?: ConfigGeneralSettingsTrustedProxyRangesList | null;
   /** Control access to the Proxy UI */
   ui_access_mode?: ConfigGeneralSettingsUiAccessMode | (string & {}) | null;
@@ -836,6 +1006,8 @@ export interface ConfigGeneralSettings {
   use_google_kms?: boolean | null;
   /** If True and LiteLLM_SpendLogs has been converted to a range-partitioned table (db_scripts/partition_spend_logs.sql), retention cleanup drops expired partitions instead of deleting rows, and pre-creates upcoming partitions. Default is False. */
   use_spend_logs_partitioning?: boolean | null;
+  /** max number of entries (virtual keys, teams, users, end users, memberships, ...) each worker keeps in its in-memory auth cache. Defaults to 200. Raise this if you have more active keys than that or auth lookups keep hitting the DB */
+  user_api_key_cache_max_size?: number | null;
   user_header_mappings?: ConfigGeneralSettingsUserHeaderMappingsList | null;
   /** [DEPRECATED] Use 'user_header_mappings' instead. When set, the header value is treated as the end user id unless overridden by user_header_mappings. */
   user_header_name?: string | null;
@@ -851,6 +1023,7 @@ export interface ConfigGeneralSettings {
 }
 export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
   S.Struct({
+    admission_queue_timeout_seconds: S.optional(S.Number),
     alert_to_webhook_url: S.optional(
       S.NullOr(ConfigGeneralSettingsAlertToWebhookUrlMap),
     ),
@@ -859,6 +1032,10 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     alerting_args: S.optional(S.NullOr(ConfigGeneralSettingsAlertingArgsMap)),
     alerting_threshold: S.optional(S.NullOr(S.Number)),
     allow_cli_sso_verification_uri_complete: S.optional(S.NullOr(S.Boolean)),
+    allow_unmanaged_response_ids: S.optional(S.NullOr(S.Boolean)),
+    allowed_file_extensions: S.optional(
+      S.NullOr(ConfigGeneralSettingsAllowedFileExtensionsList),
+    ),
     allowed_routes: S.optional(
       S.NullOr(ConfigGeneralSettingsAllowedRoutesList),
     ),
@@ -867,7 +1044,13 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
       S.NullOr(ConfigGeneralSettingsBackgroundHealthCheckModelGroupsList),
     ),
     background_health_checks: S.optional(S.NullOr(S.Boolean)),
+    blocked_file_extensions: S.optional(
+      S.NullOr(ConfigGeneralSettingsBlockedFileExtensionsList),
+    ),
     cancel_on_disconnect: S.optional(S.NullOr(S.Boolean)),
+    claude_code_gateway_managed_settings: S.optional(
+      S.NullOr(ConfigGeneralSettingsClaudeCodeGatewayManagedSettingsMap),
+    ),
     completion_model: S.optional(S.NullOr(S.String)),
     control_plane_url: S.optional(S.NullOr(S.String)),
     coordination_redis: S.optional(S.NullOr(CoordinationRedisParams)),
@@ -880,13 +1063,21 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     database_extra_connection_params: S.optional(
       S.NullOr(ConfigGeneralSettingsDatabaseExtraConnectionParamsMap),
     ),
+    database_max_idle_connection_lifetime: S.optional(S.NullOr(S.Number)),
     database_socket_timeout: S.optional(S.NullOr(S.Number)),
     database_type: S.optional(S.NullOr(S.String)),
     database_url: S.optional(S.NullOr(S.String)),
     disable_auto_add_proxy_admin_to_teams: S.optional(S.NullOr(S.Boolean)),
     disable_budget_reservation: S.optional(S.NullOr(S.Boolean)),
+    disable_env_credential_login: S.optional(S.NullOr(S.Boolean)),
+    disable_password_login_when_sso_enabled: S.optional(S.NullOr(S.Boolean)),
+    disable_responses_id_security: S.optional(S.NullOr(S.Boolean)),
+    enable_claude_code_gateway: S.optional(S.NullOr(S.Boolean)),
+    enable_openai_websocket_passthrough: S.optional(S.NullOr(S.Boolean)),
     enable_public_model_hub: S.optional(S.Boolean),
     enforce_fallback_model_access: S.optional(S.NullOr(S.Boolean)),
+    failed_login_block_seconds: S.optional(S.NullOr(S.Number)),
+    failed_login_window_seconds: S.optional(S.NullOr(S.Number)),
     forward_client_headers_to_llm_api: S.optional(S.NullOr(S.Boolean)),
     global_max_parallel_requests: S.optional(S.NullOr(S.Number)),
     health_check_concurrency: S.optional(S.NullOr(S.Number)),
@@ -896,7 +1087,16 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     key_management_system: S.optional(S.NullOr(KeyManagementSystem)),
     master_key: S.optional(S.NullOr(S.String)),
     max_batch_file_size_mb: S.optional(S.NullOr(S.Number)),
+    max_failed_login_attempts_per_source: S.optional(S.NullOr(S.Number)),
+    max_failed_login_attempts_per_source_overrides: S.optional(
+      S.NullOr(
+        ConfigGeneralSettingsMaxFailedLoginAttemptsPerSourceOverridesMap,
+      ),
+    ),
+    max_file_size_mb: S.optional(S.NullOr(S.Number)),
+    max_in_flight_requests_per_worker: S.optional(S.NullOr(S.Number)),
     max_parallel_requests: S.optional(S.NullOr(S.Number)),
+    max_queued_requests_per_worker: S.optional(S.NullOr(S.Number)),
     max_request_size_mb: S.optional(S.NullOr(S.Number)),
     max_response_size_mb: S.optional(S.NullOr(S.Number)),
     maximum_autorouter_session_retention_period: S.optional(S.NullOr(S.String)),
@@ -906,6 +1106,10 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     maximum_spend_logs_cleanup_max_batches: S.optional(S.NullOr(S.Number)),
     maximum_spend_logs_cleanup_run_budget: S.optional(S.NullOr(S.String)),
     maximum_spend_logs_retention_period: S.optional(S.NullOr(S.String)),
+    mcp_allowed_clients: S.optional(
+      S.NullOr(ConfigGeneralSettingsMcpAllowedClientsList),
+    ),
+    mcp_client_id_header: S.optional(S.NullOr(S.String)),
     mcp_internal_ip_ranges: S.optional(
       S.NullOr(ConfigGeneralSettingsMcpInternalIpRangesList),
     ),
@@ -916,12 +1120,20 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
       S.NullOr(ConfigGeneralSettingsMcpTrustedProxyRangesList),
     ),
     mcp_xff_num_trusted_hops: S.optional(S.NullOr(S.Number)),
+    missing_session_id: S.optional(
+      S.NullOr(ConfigGeneralSettingsMissingSessionId),
+    ),
     model_list_healthy_only: S.optional(S.NullOr(S.Boolean)),
     otel: S.optional(S.NullOr(S.Boolean)),
     pass_through_endpoints: S.optional(
       S.NullOr(ConfigGeneralSettingsPassThroughEndpointsList),
     ),
     pass_through_request_timeout: S.optional(S.NullOr(S.Number)),
+    password_policy_min_length: S.optional(S.NullOr(S.Number)),
+    password_policy_require_lowercase: S.optional(S.NullOr(S.Boolean)),
+    password_policy_require_numbers: S.optional(S.NullOr(S.Boolean)),
+    password_policy_require_special_characters: S.optional(S.NullOr(S.Boolean)),
+    password_policy_require_uppercase: S.optional(S.NullOr(S.Boolean)),
     plugins: S.optional(S.NullOr(ConfigGeneralSettingsPluginsList)),
     provider_url_destination_allowed_hosts: S.optional(
       S.NullOr(ConfigGeneralSettingsProviderUrlDestinationAllowedHostsList),
@@ -934,6 +1146,9 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     supported_db_objects: S.optional(
       S.NullOr(ConfigGeneralSettingsSupportedDbObjectsList),
     ),
+    transcribe_media_buckets: S.optional(
+      S.NullOr(ConfigGeneralSettingsTranscribeMediaBucketsList),
+    ),
     trusted_proxy_ranges: S.optional(
       S.NullOr(ConfigGeneralSettingsTrustedProxyRangesList),
     ),
@@ -941,6 +1156,7 @@ export const ConfigGeneralSettings = /*@__PURE__*/ S.suspend(() =>
     use_azure_key_vault: S.optional(S.NullOr(S.Boolean)),
     use_google_kms: S.optional(S.NullOr(S.Boolean)),
     use_spend_logs_partitioning: S.optional(S.NullOr(S.Boolean)),
+    user_api_key_cache_max_size: S.optional(S.NullOr(S.Number)),
     user_header_mappings: S.optional(
       S.NullOr(ConfigGeneralSettingsUserHeaderMappingsList),
     ),
@@ -1105,8 +1321,10 @@ export interface RetryPolicy {
   AuthenticationErrorRetries?: number | null;
   BadRequestErrorRetries?: number | null;
   ContentPolicyViolationErrorRetries?: number | null;
+  DefaultRetries?: number | null;
   InternalServerErrorRetries?: number | null;
   RateLimitErrorRetries?: number | null;
+  ServiceUnavailableErrorRetries?: number | null;
   TimeoutErrorRetries?: number | null;
 }
 export const RetryPolicy = /*@__PURE__*/ S.suspend(() =>
@@ -1114,8 +1332,10 @@ export const RetryPolicy = /*@__PURE__*/ S.suspend(() =>
     AuthenticationErrorRetries: S.optional(S.NullOr(S.Number)),
     BadRequestErrorRetries: S.optional(S.NullOr(S.Number)),
     ContentPolicyViolationErrorRetries: S.optional(S.NullOr(S.Number)),
+    DefaultRetries: S.optional(S.NullOr(S.Number)),
     InternalServerErrorRetries: S.optional(S.NullOr(S.Number)),
     RateLimitErrorRetries: S.optional(S.NullOr(S.Number)),
+    ServiceUnavailableErrorRetries: S.optional(S.NullOr(S.Number)),
     TimeoutErrorRetries: S.optional(S.NullOr(S.Number)),
   }),
 ).annotate({ identifier: "RetryPolicy" }) as any as S.Schema<RetryPolicy>;
@@ -1129,6 +1349,25 @@ export const UpdateRouterConfigModelGroupRetryPolicyMap =
     RetryPolicy,
   ) as any as S.Schema<UpdateRouterConfigModelGroupRetryPolicyMap>;
 
+export type UpdateRouterConfigOptionalPreCallChecksItem =
+  | "prompt_caching"
+  | "router_budget_limiting"
+  | "responses_api_deployment_check"
+  | "deployment_affinity"
+  | "session_affinity"
+  | "forward_client_headers_by_model_group"
+  | "enforce_model_rate_limits"
+  | "encrypted_content_affinity";
+export const UpdateRouterConfigOptionalPreCallChecksItem = S.String;
+
+export type UpdateRouterConfigOptionalPreCallChecksList = Array<
+  UpdateRouterConfigOptionalPreCallChecksItem | (string & {})
+>;
+export const UpdateRouterConfigOptionalPreCallChecksList =
+  /*@__PURE__*/ S.Array(
+    UpdateRouterConfigOptionalPreCallChecksItem,
+  ) as any as S.Schema<UpdateRouterConfigOptionalPreCallChecksList>;
+
 export type RoutingGroupModelsList = Array<string>;
 export const RoutingGroupModelsList = /*@__PURE__*/ S.Array(
   S.String,
@@ -1185,6 +1424,7 @@ export interface UpdateRouterConfig {
   model_group_alias?: UpdateRouterConfigModelGroupAliasMap | null;
   model_group_retry_policy?: UpdateRouterConfigModelGroupRetryPolicyMap | null;
   num_retries?: number | null;
+  optional_pre_call_checks?: UpdateRouterConfigOptionalPreCallChecksList | null;
   retry_after?: number | null;
   retry_policy?: RetryPolicy | null;
   routing_groups?: UpdateRouterConfigRoutingGroupsList | null;
@@ -1192,6 +1432,7 @@ export interface UpdateRouterConfig {
   routing_strategy_args?: UpdateRouterConfigRoutingStrategyArgsMap | null;
   tag_routing_prefix?: string | null;
   timeout?: number | null;
+  weights?: unknown | null;
 }
 export const UpdateRouterConfig = /*@__PURE__*/ S.suspend(() =>
   S.Struct({
@@ -1213,6 +1454,9 @@ export const UpdateRouterConfig = /*@__PURE__*/ S.suspend(() =>
       S.NullOr(UpdateRouterConfigModelGroupRetryPolicyMap),
     ),
     num_retries: S.optional(S.NullOr(S.Number)),
+    optional_pre_call_checks: S.optional(
+      S.NullOr(UpdateRouterConfigOptionalPreCallChecksList),
+    ),
     retry_after: S.optional(S.NullOr(S.Number)),
     retry_policy: S.optional(S.NullOr(RetryPolicy)),
     routing_groups: S.optional(S.NullOr(UpdateRouterConfigRoutingGroupsList)),
@@ -1222,6 +1466,7 @@ export const UpdateRouterConfig = /*@__PURE__*/ S.suspend(() =>
     ),
     tag_routing_prefix: S.optional(S.NullOr(S.String)),
     timeout: S.optional(S.NullOr(S.Number)),
+    weights: S.optional(S.NullOr(S.Unknown)),
   }),
 ).annotate({
   identifier: "UpdateRouterConfig",