MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 11 callersMethodpost_call
(self, original_response, input=None, api_key=None, additional_args=None)
litellm/types/utils.py:4885
↓ 11 callersMethodpost_mcp_call_hook
Run guardrails configured for ``post_mcp_call`` against an MCP tool result. The MCP counterpart of ``post_call_success_hook``: guard
litellm/proxy/utils.py:4612
↓ 11 callersFunctionpostgres_span_names
The ``postgres.{verb} {table}`` names OTel would render for every DB service event the code under test emits, in emission order, once the hook tas
tests/unit/proxy/conftest.py:498
↓ 11 callersMethodpre_call_check
Synchronous pre-call check for model rate limits. Raises RateLimitError if deployment exceeds TPM/RPM limits.
litellm/router_utils/pre_call_checks/model_rate_limit_check.py:302
↓ 11 callersFunctionprint_verbose
( print_statement, logger_only: bool = False, log_level: Literal["DEBUG", "INFO", "ERROR"] = "DEBU
litellm/utils.py:593
↓ 11 callersFunctionpromoted_baggage
Identity values to write into Baggage, filtered to ``promoted_keys``. ``promoted_keys`` selects from ``_PROMOTABLE``; ``metadata_keys`` selects
litellm/integrations/otel/model/baggage.py:69
↓ 11 callersFunctionprovider_edge_base
The api_base an edge-wired deployment should register with, using this process's fixture-mode and edge-host configuration: None in live mode, the
tests/e2e/e2e_config.py:265
↓ 11 callersFunctionproxy_exception_from_http_exception
(exc: StarletteHTTPException, headers: dict[str, str])
litellm/proxy/common_request_processing.py:849
↓ 11 callersFunctionproxy_path
(route: Route, marker: str)
tests/integration/providers/_assistants_route_errors_support.py:86
↓ 11 callersFunctionq
litellm/proxy/_experimental/out/_next/static/chunks/2_9r2jeem0dh8.js:1
↓ 11 callersFunctionr
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/0ayyip3c-drm1.js:1
↓ 11 callersFunctionraise_identity_failure
(failure: AgentIdentityFailure, status_code: int = 403)
litellm/proxy/agent_endpoints/managed_identity.py:68
↓ 11 callersMethodraise_passthrough_exception
Raise a passthrough exception for guardrail violations. This helper method should be used by guardrails when they detect a violation
litellm/integrations/custom_guardrail.py:520
↓ 11 callersFunctionrate
(provider: str, tier_field: str, model: str = MODEL)
tests/integration/providers/_dashscope_video_support.py:121
↓ 11 callersFunctionraw_chat
(gateway: Gateway, group: str, marker: str, **extra: JsonValue)
tests/integration/routing/test_payment_required_cooldown.py:171
↓ 11 callersMethodread_body_back_everywhere
GET `path` on every replica that serves it, polling each to poll_timeout until `settled` accepts its body, and fail naming the first replica t
tests/e2e/proxy_client.py:839
↓ 11 callersFunctionread_pins
(redis: Redis)
tests/integration/routing/_prompt_cache_pin_support.py:85
↓ 11 callersMethodread_secret
( &'a self, _name: &'a str, _settings: &'a KeyManagementSettings, _environment
litellm-rust/crates/secrets/tests/resolution.rs:183
↓ 11 callersMethodrecord
( self, *, payload: Mapping[str, object], request_status: Literal["success", "
litellm/proxy/db/request_error_tracking.py:93
↓ 11 callersMethodrecord
( self, kwargs: _MetricKwargs, response_obj: _ResponseView | None, start_time:
litellm/integrations/otel/plumbing/metrics.py:233
↓ 11 callersMethodrecord
Record a terminal outcome, allowing teardown errors to replace a pass.
tests/rust-python-harness/shared/reporting/models.py:110
↓ 11 callersMethodrecreate_prisma_client
Recreate both writer and reader Prisma clients. The writer reconnect path in PrismaClient calls `self.db.recreate_prisma_client(...)`
litellm/proxy/db/routing_prisma_wrapper.py:248
↓ 11 callersFunctionredact_nested_match_and_regex_keys
Deep-copy `payload` and replace every configured string field with "[REDACTED]" anywhere in nested dict/list structures. Used for guardr
litellm/litellm_core_utils/core_helpers.py:990
↓ 11 callersFunctionreencrypt_stored_values
(database: SupportsRawQueries, *, from_key: str, to_key: str)
litellm/proxy/db/master_key_migration.py:96
↓ 11 callersFunctionrefresh_proxy_server_request_body_snapshot
Re-snapshot ``data["proxy_server_request"]["body"]`` from the current state of ``data``. ``add_litellm_data_to_request`` takes the initial s
litellm/proxy/litellm_pre_call_utils.py:2577
↓ 11 callersMethodrelease_refresh_from_the_loop
(self)
tests/unit/llms/bedrock/event_loop_probe.py:55
↓ 11 callersFunctionreplace_model_in_jsonl
(file_content: FileTypes, new_model_name: str, keep_line_numbers: bool = False)
litellm/litellm_core_utils/batch_utils.py:169
↓ 11 callersFunctionreported_cost
(payload: object)
litellm/llms/edenai/common_utils.py:51
↓ 11 callersFunctionrequest
( method: str = "post", path: str = "/chat/completions", *, headers: dict[str, str] | None = N
tests/e2e_harness/test_fixture_canonical.py:19
↓ 11 callersFunctionrequest_root_span
The anchored request root span, or ``None`` outside a proxy request.
litellm/integrations/otel/plumbing/context.py:78
↓ 11 callersFunctionrequire_proxy_client
Return the shared ``ProxyClient`` plus the master key the HTTP probes authenticate with, or hard-fail the test. Both planes of the built ``Pr
tests/e2e/claude_code/_env.py:84
↓ 11 callersMethodresolve
( &self, reference: &str, environment: &(dyn Lookup + Send + Sync), )
litellm-rust/crates/secrets/src/oidc.rs:145
↓ 11 callersMethodresolve_agent_id
( jwt_handler: JWTHandler, jwt_valid_token: Mapping[str, object], agent_registry: Agen
litellm/proxy/auth/handle_jwt.py:1579
↓ 11 callersFunctionresolve_dashscope_endpoints
Derive DashScope's protocol-specific bases from any standard DashScope base. Accepts the domain root, an OpenAI-compatible base, an Anthropi
litellm/llms/dashscope/common_utils.py:79
↓ 11 callersFunctionresolve_dashscope_openai_api_base
( api_base: str | None, env_var: str = "DASHSCOPE_API_BASE", default: str = f"{DASHSCOPE_DEFAULT_O
litellm/llms/dashscope/common_utils.py:119
↓ 11 callersFunctionresolve_mcp_auth
Resolve the auth value for an MCP server. Priority: 1. ``mcp_auth_header`` — per-request/per-user override 2. OAuth2 client_credentials t
litellm/proxy/mcp_server/oauth2_token_cache.py:330
↓ 11 callersFunctionresolve_session_bearer
Classify an ``Authorization`` value presented at the aggregate MCP edge. Strips an optional ``Bearer`` scheme, then returns ``NotSessionBearer``
litellm/proxy/mcp_server/outbound_credentials/session_credentials.py:223
↓ 11 callersFunctionresolve_skill_visibility
``skill_visibility`` with the access-group grants of the key and its team folded in.
litellm/proxy/anthropic_endpoints/claude_code_endpoints/claude_code_skill_access.py:159
↓ 11 callersFunctionresourceLiteLLMCredential
()
terraform/provider/litellm/resource_credential.go:8
↓ 11 callersMethodrespond
(self, request: Request)
tests/integration/providers/test_malformed_tools_chaos.py:127
↓ 11 callersMethodrespond
(request: httpx.Request)
tests/unit/llms/anthropic/test_anthropic_common_utils.py:3795
↓ 11 callersFunctionresponse_format_schema
(value: object)
litellm/llms/base_llm/base_utils.py:278
↓ 11 callersFunctionresponses_request
(model: str, **extra: JsonValue)
tests/integration/streaming/faked_responses_case.py:187
↓ 11 callersMethodresult
(self, pending: NativePending[_T])
litellm/redis/stores/_route.py:54
↓ 11 callersMethodretain
(self, event: Mapping[str, object])
litellm/llms/openai/live/handler.py:256
↓ 11 callersFunctionretry_upstream_rate_limit
( fetch: Callable[[], Awaitable[T]], server_name: str, *, max_retries: int = MCP_UPSTREAM_RATE
litellm/proxy/mcp_server/faults/list_outcomes.py:264
↓ 11 callersMethodrun
(self, code: str, /)
litellm/llms/litellm_proxy/skills/sandbox_executor.py:31
↓ 11 callersMethodrun
(self, context)
tests/unit/router_strategy/test_complexity_router.py:7601
↓ 11 callersFunctionrun_turn
( handler: dh.DeepAgentsHandler, ctx: SessionContext, prompt: str, approve: bool = True, )
tests/unit/harness/handlers/test_deepagents_handler.py:87
↓ 11 callersFunctionrun_v1_migrations
( database_url: str | None, tmp_path: Path, extra_env: Mapping[str, str], test_name: str,
tests/integration/database/test_v1_migration_error_logging.py:160
↓ 11 callersFunctionrust_routed
(headers: Mapping[str, str])
tests/integration/providers/_anthropic_safeguards_beta_support.py:218
↓ 11 callersFunctions
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/2aatum0yo4cv8.js:1
↓ 11 callersMethodschedule
Queue an evicted client for closing once it is idle and out of grace. Past ``max_pending`` the client is left to the collector instead, so a
litellm/caching/evicted_client_closer.py:223
↓ 11 callersFunctionscripted_message
( identity: str, model: str = "databricks-claude-sonnet-4-6", **extra: JsonValue )
tests/integration/messages_endpoint/providers/databricks/test_databricks_native_messages_wire.py:194
↓ 11 callersFunctionsearch_provider_cost_per_query
Calculate cost for search-only providers. Returns (input_cost, output_cost) where input_cost = queries * cost_per_query Supports tiered
litellm/search/cost_calculator.py:62
↓ 11 callersFunctionsearch_skills
( query: str, skills: Sequence[LiteLLM_SkillsTable], top_k: int, router: Router | None, em
litellm/llms/litellm_proxy/skills/skill_search.py:124
↓ 11 callersMethodselect
( self, model: str, request_kwargs: dict[str, object] | None = None, # mutable-ok: th
litellm/router_core.py:36
↓ 11 callersMethodsend
(self, pod: OwnedProxy, call: Call)
tests/integration/observability/test_otel_team_sampling_across_pods.py:185
↓ 11 callersMethodsend
( self, endpoint: str, marker: str, fields: Mapping[str, JsonValue] = MappingP
tests/integration/spend/test_spend_log_caller_session.py:206
↓ 11 callersMethodsend_text
(self, data: str)
litellm/litellm_core_utils/realtime_streaming.py:152
↓ 11 callersFunctionsent_effort
(request: Request)
tests/integration/routing/_reasoning_effort_levels_support.py:283
↓ 11 callersMethodserve
(self, path: str, served: _Served)
tests/unit/litellm_core_utils/test_image_handling.py:736
↓ 11 callersFunctionsession_stats_snapshot
Read-only copy of the per-session VCR stats. Used by the summary.
tests/_vcr_conftest_common.py:1595
↓ 11 callersMethodset
``set_optional_pre_call_checks``: drop the toggleable checks the list leaves out, then add it.
litellm/router_rust/steps/callbacks.py:325
↓ 11 callersMethodset_cache
(self, key, value, **kwargs)
tests/unit/proxy/common_utils/test_user_api_key_cache.py:35
↓ 11 callersMethodset_llm_deployment_success_metrics
( self, request_kwargs: dict, start_time, end_time, enum_values: UserA
litellm/integrations/prometheus.py:3506
↓ 11 callersFunctionset_response_cost_in_hidden_params
Record a provider-reported cost where the cost calculator looks before the price map.
litellm/litellm_core_utils/core_helpers.py:1165
↓ 11 callersMethodset_session_routing
Store a routing override for a session. Called by guardrails when they detect sensitive data and want to route the session t
litellm/proxy/hooks/sensitive_data_routing.py:84
↓ 11 callersMethodsettled
(self)
litellm-rust/crates/redis-client/src/batch.rs:131
↓ 11 callersFunctionsetup
( call_type: str, args: tuple[object, ...], kwargs: Mapping[str, object], start_time: datetime
litellm/rust_bridge/host/callbacks_legacy_python.py:50
↓ 11 callersMethodsetup_gdpr_guardrail
Setup guardrail with all GDPR patterns (mimics the policy template)
tests/unit/proxy/guardrails/guardrail_hooks/content_filter/test_gdpr_policy_e2e.py:22
↓ 11 callersFunctionshould_convert_tool_call_to_json_mode
Determine if tool calls should be converted to JSON mode
litellm/litellm_core_utils/llm_response_utils/convert_dict_to_response.py:715
↓ 11 callersMethodshould_retry_this_error
``True`` when the error is retried in the same group; otherwise the error is raised.
litellm/router_rust/surface/retries.py:246
↓ 11 callersFunctionskill_search_text
(skill: LiteLLM_SkillsTable)
litellm/llms/litellm_proxy/skills/skill_search.py:77
↓ 11 callersMethodskip_trivia
Whitespace, comments, and backslash continuations; newlines too when `newlines`.
litellm-rust/crates/python-compat/src/literal.rs:103
↓ 11 callersMethodsource
(&self)
litellm-rust/crates/llms-types/src/formats/ocr/request.rs:21
↓ 11 callersMethodspan
(s)
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 11 callersFunctionsplit_bedrock_region_path
Split a ``<region>/<model-id>`` routing path into the region and the id AWS receives. ``bedrock/us-gov-west-1/openai.gpt-oss-20b-1:0`` -> ``("us-
litellm/llms/bedrock/common_utils.py:1235
↓ 11 callersMethodstart
(self, pool: ThreadPoolExecutor, fleet: Sequence[Team])
tests/integration/management/test_team_delete_chaos.py:226
↓ 11 callersFunctionstarted
(ctx: SessionContext)
tests/unit/harness/handlers/test_deepagents_handler.py:81
↓ 11 callersFunctionstored_manifest
(name: str)
tests/integration/management/test_claude_code_plugin_installation_preference.py:118
↓ 11 callersMethodstrip_reasoning_the_targets_cannot_decrypt
Drop the encrypted reasoning that none of ``target_deployments`` minted or shares an encryption boundary with, keeping each item's re
litellm/router_utils/pre_call_checks/encrypted_content_affinity_check.py:439
↓ 11 callersMethodsubmit
(self, fn, *args, **kwargs)
tests/unit/passthrough/test_streaming_interrupt_spend_tracking.py:33
↓ 11 callersMethodsync_policies_from_db
Sync policies from the database to in-memory registry. - Production versions are loaded into _policies (by policy name) for resolutio
litellm/proxy/policy_engine/policy_registry.py:650
↓ 11 callersMethodsync_read_secret
(self, secret_name: str, primary_secret_name: str | None = None)
tests/unit/rust_bridge/host/test_secret_manager.py:160
↓ 11 callersFunctiont
litellm/proxy/_experimental/out/_next/static/chunks/1a2lzi3ursud4.js:1
↓ 11 callersFunctionteam_alias
(candidate: Gateway, team: str)
tests/integration/management/test_scim_group_pathless_patch.py:54
↓ 11 callersFunctionteam_key
(candidate: Gateway, team: str, user: str)
tests/integration/_support/scim.py:125
↓ 11 callersFunctionteam_member_budget_entity_id
(user_id: str, team_id: str)
litellm/proxy/hooks/model_max_budget_limiter.py:110
↓ 11 callersFunctionteam_member_update
[BETA] Update team member budgets and team member role
litellm/proxy/management_endpoints/team_endpoints.py:4178
↓ 11 callersMethodtell
(self)
litellm/llms/black_forest_labs/image_edit/transformation.py:64
↓ 11 callersFunctiontext_block
(text: str)
tests/integration/_support/safeguards_rig.py:645
↓ 11 callersFunctionthrowError
(s,o)
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 11 callersFunctiontimeout
(model: str = "gpt-4o")
tests/unit/test_router/core/scripted_host.py:147
↓ 11 callersMethodtoJSON
()
litellm/proxy/_experimental/out/_next/static/chunks/085r5ieqhajig.js:1
↓ 11 callersMethodtoString
()
litellm/proxy/_experimental/out/_next/static/chunks/085r5ieqhajig.js:1
↓ 11 callersFunctionto_string
(s)
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 11 callersMethodtoken
(self, index: int = 0)
tests/integration/observability/test_guardrail_effects.py:2423
↓ 11 callersFunctiontoken_counter
The same as `litellm.litellm_core_utils.token_counter`. Kept for backwards compatibility.
litellm/utils.py:3158
↓ 11 callersFunctiontool_use_label
(city: str)
tests/integration/providers/_cache_control_marks_support.py:69
← previousnext →6,401–6,500 of 202,479, ranked by callers