MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 6 callersFunctionrewrite_streamed_response_ids
Record ownership of the response object streamed back by a Responses API passthrough and swap its managed id into every SSE frame, so a strea
litellm/proxy/pass_through_endpoints/managed_id_rewriter.py:903
↓ 6 callersFunctionrole_out_of_guardrail_scope
( role: str, *, skip_system_message: bool, skip_tool_message: bool, scan_only_tool_results
litellm/llms/base_llm/guardrail_translation/utils.py:284
↓ 6 callersFunctionrotate_search_tools_master_key
Re-encrypt the litellm_params values that decrypt under the current key with the key in force after rotation (LITELLM_SALT_KEY when set, otherwise
litellm/proxy/search_endpoints/search_tool_registry.py:143
↓ 6 callersFunctionround_trip
(socket: &mut litellm_http::websocket::UpstreamWebSocket)
litellm-rust/crates/http/tests/websocket.rs:213
↓ 6 callersMethodroute
(self, team: str | None, request: str, filtered: Sequence[str] | None)
litellm/router_rust/surface/indexes.py:470
↓ 6 callersFunctionroute_failure
Maps a route failure once. A core decline and a Python exception raised inside the call are returned as they are, everything else is mapped and handed
litellm-rust/crates/python-bridge/src/errors.rs:79
↓ 6 callersFunctionroute_failure_proxy_exception
( exc: Exception, message: str, default_status_code: int, headers: dict[str, str] | None = Non
litellm/proxy/common_utils/openai_error_payload.py:195
↓ 6 callersFunctionrouter_identity
Stable identity for a complexity-router deployment, across tuning edits.
litellm/router_strategy/auto_router/auto_router_tuning_baseline.py:76
↓ 6 callersFunctionrouter_init_args
The arguments ``Router.__init__`` accepts (was ``Router.get_valid_args``). Needs no router instance.
litellm/router_api.py:280
↓ 6 callersMethodrouting_group_named
The callable routing group named `name`: a model group of that name shadows it.
litellm-rust/crates/router-core/src/registry.rs:999
↓ 6 callersMethodrouting_strategy_init
``routing_strategy_init``: the default selector only. As on the Python core the router's ``routing_strategy`` setting stays, so picks keep fol
litellm/router_rust/surface/legacy_members.py:281
↓ 6 callersFunctionrow_key
(request_id: str)
tests/integration/providers/_mantle_gpt_prompt_cache_support.py:474
↓ 6 callersMethodrow_view
(&self, row: &'r Row)
litellm-rust/crates/router-core/src/lookup.rs:158
↓ 6 callersMethodrows_for
(self, model: str)
tests/integration/_support/prompt_cache_breakpoint.py:172
↓ 6 callersFunctionrows_to_py
Lookup rows as the host reads them.
litellm-rust/crates/python-bridge/src/router/mod.rs:1115
↓ 6 callersFunctionrun
(marker: str)
tests/unit/litellm_core_utils/test_litellm_logging.py:10727
↓ 6 callersFunctionrun_composed
( py: Python<'_>, machine: MachineFactory, binding: SyntheticBinding, hooks: SyntheticHooks,
litellm-rust/crates/host-python/tests/support/mod.rs:524
↓ 6 callersMethodrun_migration_job
The migration job's whole run: `setup_database`, then the request-log indexes, built synchronously so the job exits only once they are in plac
litellm-proxy-extras/litellm_proxy_extras/utils.py:1509
↓ 6 callersFunctionrun_off_loop_if_fips
Run a password KDF inline, or on the single KDF thread in FIPS mode.
litellm/proxy/utils.py:8541
↓ 6 callersFunctionrun_python_tests
( selectors: Sequence[str], repo_root: Path, backend: Backend, spec: BackendSpec, pytest_a
tests/rust-python-harness/shared/unit_runners/python_runner.py:81
↓ 6 callersFunctionrun_scenario
( cb: CallbackRig, scenario: Scenario, who: Caller, secret: Canary, endpoint: str, out
tests/integration/security/test_callback_credentials.py:152
↓ 6 callersFunctionrun_sync
( py: Python<'_>, future: F, map_error: fn(E) -> PyErr, )
litellm-rust/crates/host-python/src/runtime.rs:98
↓ 6 callersFunctionrun_sync_value
(py: Python<'_>, future: F)
litellm-rust/crates/python-bridge/src/execution.rs:32
↓ 6 callersFunctionruntime
()
litellm-rust/crates/host-python/src/runtime.rs:66
↓ 6 callersFunctionrust_core
A ``RustCore`` on ``host``, built the way ``make_core`` builds one.
tests/unit/test_router/core/scripted_host.py:489
↓ 6 callersFunctionrz
(e)
litellm/proxy/_experimental/out/_next/static/chunks/2aatum0yo4cv8.js:1
↓ 6 callersFunctions
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/14v3irtjifjt2.js:1
↓ 6 callersFunctionsafeBuildUrl
(s,o,{selectedServer:i=""}={})
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 6 callersFunctionsafe_divide_seconds
Safely divide seconds by denominator, handling zero division. Args: seconds: Time duration in seconds denominator: The divis
litellm/litellm_core_utils/core_helpers.py:83
↓ 6 callersFunctionsafe_json_structure
Rebuild data out of JSON-native pieces while detecting circular references. If a circular reference is detected then a marker string is retur
litellm/litellm_core_utils/safe_json_dumps.py:28
↓ 6 callersFunctionsame_header
Whether two HTTP header names are the same one. They are case-insensitive (RFC 7230 3.2).
litellm/types/mcp.py:342
↓ 6 callersFunctionsatisfied
(states: Sequence[_State])
tests/unit/test_router/test_router_callback_hook_sequence.py:241
↓ 6 callersFunctionsave_background_health_checks_to_db
Save background health check results to database for each model. Maps health check endpoints back to their original models to get model_name
litellm/proxy/health_endpoints/_health_endpoints.py:986
↓ 6 callersFunctionscenario_api_base
(scenario: Scenario, scenario_id: str, response: StoredResponse)
tests/integration/routing/test_fallback_hop_foreign_encrypted_reasoning.py:297
↓ 6 callersMethodschedule_stream_failure_logging
( litellm_logging_obj: LiteLLMLoggingObj, endpoint_type: EndpointType, request_body: d
litellm/proxy/pass_through_endpoints/streaming_handler.py:75
↓ 6 callersFunctionscim_data
(metadata: Mapping[str, JsonValue])
tests/integration/management/test_scim_group_pathless_patch.py:75
↓ 6 callersFunctionscope
(provider: Option<&str>, model_id: Option<&str>, affinity: bool)
litellm-rust/crates/inference-responses/src/ids.rs:287
↓ 6 callersFunctionscope_key
``main``'s ``f"{{{key}:{value}}}:{suffix}"``.
tests/unit/redis/stores/fake_rate_limits.py:39
↓ 6 callersMethodscript
( self, source: str, run: RegisteredScript, keys: Sequence[str], args: Sequence[_ScriptArg] )
litellm/caching/redis_batch.py:327
↓ 6 callersFunctionsdk_settled
(call: Callable[[], T])
tests/integration/routing/test_priority_scheduler_queue_cleanup.py:273
↓ 6 callersFunctionseal_bridge_authorization_code
Seal the upstream authorization code and the SSO-captured litellm user into a gateway authorization code. The DCR client only echoes this opaque v
litellm/proxy/mcp_server/discoverable_endpoints.py:342
↓ 6 callersMethodsearch
( &self, request: Request<SearchPoints>, )
litellm-rust/crates/cache-qdrant-semantic/tests/support/mod.rs:232
↓ 6 callersMethodsearch_keys
(self, scope: DailyActivityScope, *, search: str, limit: int)
litellm/repositories/daily_activity_repository.py:201
↓ 6 callersMethodsecond
()
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 6 callersFunctionsecret_manager_would_be_consulted
Returns True if a `get_secret` read for `secret_name` would actually reach the hosted manager. Mirrors the gating `get_secret` applies below
litellm/secret_managers/main.py:514
↓ 6 callersMethodsecret_names
(self)
litellm-rust/crates/inference-ocr/src/provider_config.rs:123
↓ 6 callersMethodsecret_names
(&self)
litellm-rust/crates/gateway/src/redis/secrets.rs:48
↓ 6 callersFunctionsecret_reader_returning
(secret: str | None)
tests/unit/llms/base_llm/auth/test_client_credentials.py:75
↓ 6 callersFunctionselectControl
(control: MountedFieldControlProps)
ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcpFieldRules.ts:26
↓ 6 callersFunctionselectTriggerControl
(control: MountedFieldControlProps)
ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcpFieldRules.ts:31
↓ 6 callersFunctionsend
(row: Row, target: ProxyTarget)
tests/integration/providers/_idle_upstream_support.py:556
↓ 6 callersMethodsend_email
( self, from_email: str, to_email: List[str], subject: str, html_body:
enterprise/litellm_enterprise/enterprise_callbacks/send_emails/base_email.py:971
↓ 6 callersMethodsend_key_created_email
Send email to user after creating key for the user
enterprise/litellm_enterprise/enterprise_callbacks/send_emails/base_email.py:129
↓ 6 callersMethodsend_message
(self, agent_id: str, key: str, body: A2AJsonRpcRequest)
tests/e2e/a2a/a2a_client.py:262
↓ 6 callersMethodsend_model_deprecation_alert
Alert on the router's deprecated and imminent models, True when one was sent The daily lock is claimed only once there is something to say, s
litellm/integrations/SlackAlerting/slack_alerting.py:1107
↓ 6 callersMethodsend_text
(self, data: str)
litellm/llms/bedrock/realtime/handler.py:125
↓ 6 callersFunctionsent_overrides
(host: ScriptedHost)
tests/unit/test_router/core/test_gaps_request_overrides.py:33
↓ 6 callersFunctionseparate_cached_messages
Returns the first continuous block of cached messages, plus every system message once any message is cached, since generateContent rejects a
litellm/llms/vertex_ai/context_caching/transformation.py:109
↓ 6 callersFunctionserved_stream_output_texts
(chunks: Sequence[object])
litellm/litellm_core_utils/served_output_texts.py:124
↓ 6 callersMethodservice_success_hook
(self, payload: ServiceLoggerPayload)
litellm/integrations/prometheus_services.py:245
↓ 6 callersMethodsession_configuration_request
``` { "model": string, "generationConfig": { "candidateCount": integer, "max
litellm/llms/gemini/realtime/transformation.py:1870
↓ 6 callersFunctionsession_for
(system, first)
tests/unit/proxy/guardrails/guardrail_hooks/test_straiker.py:2016
↓ 6 callersFunctionsession_public_key_pem
The PEM public half of the current RS256 signing key: the only material a downstream validator (an external gateway verifying ``kid``-matched toke
litellm/proxy/mcp_server/outbound_credentials/session_token.py:221
↓ 6 callersMethodset
A spend reset: memory, then Redis, each for `ttl` or its own default.
litellm-rust/crates/redis-store/src/spend_counters/mod.rs:305
↓ 6 callersFunctionset_extra_attribute
(model: BaseModel, name: str, value: object)
litellm/litellm_core_utils/hidden_params.py:87
↓ 6 callersMethodset_local
(self, entry: RateLimitEntry, value: LocalValue, ttl_seconds: float | None)
tests/unit/redis/stores/fake_rate_limits.py:132
↓ 6 callersMethodset_local
(self, key: str, value: object, ttl_seconds: float | None = None)
tests/unit/redis/stores/auth_cache_double.py:146
↓ 6 callersMethodset_param
(self, param_name: str, param_value: object)
litellm/proxy/roi_calculator/sync.py:52
↓ 6 callersMethodset_response_timing_metrics
Keep ``_response_ms`` / ``litellm_overhead_time_ms`` for a result that has no ``_hidden_params``.
litellm/litellm_core_utils/litellm_logging.py:999
↓ 6 callersMethodset_timing_metrics
Set response timing metrics
litellm/litellm_core_utils/llm_response_utils/response_metadata.py:210
↓ 6 callersFunctionset_user_provider_credential_cache
Connect path: overwrite the key with the new connection's ciphertext so a stale set-if-absent fill from a pre-connect read cannot resurrect a
litellm/proxy/credential_endpoints/user_provider_credentials.py:254
↓ 6 callersFunctionsettings
(wire: Wire)
litellm-rust/crates/testkit/tests/configure.rs:7
↓ 6 callersMethodsettings
(&self)
litellm-rust/crates/router-core/src/lib.rs:414
↓ 6 callersFunctionsettings_path_for
(target: Target)
litellm/proxy/client/cli/commands/configure_profiles.py:66
↓ 6 callersFunctionsettle
( send: Callable[[str], httpx.Response], status: int, *, burst: int = GATEWAY_BURST, seconds: float = 60 )
tests/integration/authorization/_hidden_alias_budget.py:157
↓ 6 callersMethodsettle
Runs one hook step to its value, handing each awaitable it produces to the caller.
litellm-rust/crates/host-python/src/driver.rs:218
↓ 6 callersFunctionsetup_locks
(targets: Sequence[Target])
litellm/proxy/client/cli/commands/configure_profiles.py:112
↓ 6 callersFunctionsetup_models
(setup: HarnessSessionSetup)
tests/unit/llms/pi/harness/test_transformation.py:238
↓ 6 callersMethodsharedDepth
(e)
litellm/proxy/_experimental/out/_next/static/chunks/085r5ieqhajig.js:1
↓ 6 callersMethodshould_auto_execute_tools
Check if we should auto-execute tool calls. Auto-execution requires EVERY MCP reference to opt in with ``require_approval="never"``.
litellm/responses/mcp/litellm_proxy_mcp_handler.py:584
↓ 6 callersMethodshould_fake_stream
Returns True if the model/provider should fake stream
litellm/llms/base_llm/chat/transformation.py:169
↓ 6 callersMethodshould_filter_anthropic_beta_headers
(self)
litellm/llms/anthropic/messages/native/transformation.py:144
↓ 6 callersMethodshould_retry
( &self, snapshot: &RouterStateSnapshot, request: &RouterRequest, failure: &At
litellm-rust/crates/router-core/src/call.rs:1163
↓ 6 callersMethodshould_run_callback
(self, callback: litellm.CALLBACK_TYPES, litellm_params: object, event_hook: str)
litellm/litellm_core_utils/litellm_logging.py:2969
↓ 6 callersMethodshould_run_guardrail
(self, data: Mapping[str, object], event_type: GuardrailEventHooks)
litellm/proxy/guardrails/guardrail_hooks/llm_as_a_judge/__init__.py:230
↓ 6 callersFunctionshould_suppress_spend_log_tracebacks
Return ``True`` when spend-log traceback suppression should apply. Suppression only kicks in when both: * the operator opted in via the env
litellm/proxy/spend_tracking/spend_log_error_logger.py:46
↓ 6 callersMethodshutdown
Close every destination processor, once the spans in flight have landed. ``on_end`` runs on whichever thread ends a span and can reach this f
litellm/integrations/otel/plumbing/providers.py:939
↓ 6 callersFunctionsign_es256_jwt
Signs ``claims`` with the PEM key, stamping ``kid`` as its RFC 7638 thumbprint so a verifier can look the signing key up in the published JWKS by
litellm/llms/base_llm/auth/jwt_signing.py:105
↓ 6 callersFunctionsign_oci_request
Route to the appropriate OCI signing method based on what credentials are present. If ``oci_signer`` is in optional_params, use the OCI SDK
litellm/llms/oci/common_utils.py:496
↓ 6 callersFunctionsignal_group
(group: int, action: int)
tests/integration/_support/process.py:141
↓ 6 callersFunctionsigned_blocks
(deltas: Sequence[Mapping[str, JsonValue]])
tests/integration/_support/anthropic_thinking.py:192
↓ 6 callersFunctionskill_visibility
( user_api_key_dict: UserAPIKeyAuth | None, inherited: AccessGroupSkills = NO_ACCESS_GROUP_SKILLS )
litellm/proxy/anthropic_endpoints/claude_code_endpoints/claude_code_skill_access.py:83
↓ 6 callersFunctionsl
litellm/proxy/_experimental/out/_next/static/chunks/2k-fkubkt19cc.js:8
↓ 6 callersFunctionslice_around_compaction_block
Apply Anthropic's "drop everything before the compaction block" rule. Returns ``(sliced_messages_with_compaction_block, compaction_block_dict)``
litellm/llms/anthropic/messages/context_management/editors/compact.py:140
↓ 6 callersFunctionslow_upstream_sync_client
(body: bytes = b"")
tests/unit/llms/bedrock/slow_upstream.py:51
↓ 6 callersFunctionspeech_deployment
(scenario: Scenario, wire_url: str)
tests/integration/providers/_gemini_tts_voice_support.py:382
↓ 6 callersMethodspend_logs
(self)
tests/integration/spend/test_shutdown_flush.py:352
↓ 6 callersFunctionspend_reader
(repository: ConfigRepository)
litellm/proxy/management_endpoints/roi_calculator_endpoints.py:306
↓ 6 callersFunctionspend_reset_unit_of_work
(new_batch: Callable[[], PrismaBatch])
litellm/repositories/unit_of_work.py:191
↓ 6 callersFunctionspend_row_lands
(identity: str)
tests/integration/routing/test_priority_scheduler_queue_cleanup.py:344
← previousnext →13,501–13,600 of 202,479, ranked by callers