Post-call pipeline rewrites on buffered streams failed open on three shapes:
chat streams with n > 1 (the rebuilt response collapsed every choice into
index 0), streams that ended without a finish marker, and Responses streams
whose final event carried no response envelope.
The chat handler now rebuilds the ended stream one choice index at a time and
writes each choice's rewrite back to that choice's buffered deltas. The
Anthropic handler writes an unended stream's rewrite across its text deltas.
The Responses handler spreads an envelope-less rewrite over the buffered
output_text events, still failing open when a scanned event cannot be placed.
Tool-call rewrites on n > 1 chat streams keep failing open.
Applying a database row dropped the runtime layer for every key the row
carried, including keys the config file owns. Those runtime entries hold
the env-resolved config values, so after a reload a key written as
os.environ/<NAME> read back as that literal string. The store now keeps
the runtime entry for a key the config owns and clears only the rest.
Visible as store_model_in_db silently turning itself off: the reload read
the raw reference, coerced it to False, and overwrote the resolved global.
A write into a settings store for a key the config file declares used to
land in the runtime layer and then lose to the config on every read, so
the caller saw success while nothing changed. It now raises
ConfigOwnedKeyError, and the allowed-IP routes turn that into a 400
naming the key instead of reporting success on a list they never changed.
Both allowed-IP routes now build a new list rather than mutating the one
the config layer holds, and the os.environ resolver rebuilds the config
it is given instead of writing back into it, so a reader can no longer
corrupt the raw values the store keeps for provenance.
The database reload leaves a config-owned key alone rather than writing a
normalized copy back over it, which would now raise and abort the rest of
the reconcile pass.
retrieve_batch returned a terminal batch from the DB before checking that the key may use the model encoded in a unified batch id; the grant check now runs right after pre-call processing. The vector store file list helper authorized data["model"] through handle_model_based_routing even when the vector store registry set it server-side and even with no caller, which crashed on a None key; it now authorizes only a caller-supplied hint and resolves credentials directly.
PR #40243 started carrying the upstream error body on InternalServerError so the Responses response.failed event can report the provider's code and message, and openai's APIError.__init__ took the body's type along with it. The proxy then answered an OpenAI-compatible upstream 500 with type server_error while a 502 and a 503 kept internal_server_error, and the integration contract in test_observed_routing.py went red. Pin the type the way RateLimitError pins throttling_error, keeping the body.
Resolves handler.py against main's SearchOutcome refactor: the file is
main's version plus this PR's substantive hunks only (the RichWebSearchInput
import, the rich= wiring at the three _execute_search call sites, the
_rich_search_input and _provider_supports_rich_search helpers, and the
_execute_search forwarding), so the 88-column re-wrap noise the PR carried
is gone and the diff against main is the feature alone. RichWebSearchInput
sits beside main's new SearchSucceeded/SearchFailed types, and main's two
_execute_search test stubs accept the new rich argument.
#41515 added the cost map entry transcribe/StartTranscriptionJob under a
new litellm_provider value "transcribe" without registering that provider
anywhere else, so litellm.models_by_provider had no "transcribe" key.
test_models_by_provider derives its provider set from the cost map itself,
so it went red on main. The user-visible half is that get_provider_models
returned None for the provider, which get_known_models_from_wildcard turns
into an empty list, leaving a transcribe/* key or team grant resolving to
no models.
Mirror the aws_polly registration: an enum member, a model set, an
ingestion branch, and a models_by_provider entry. Amazon Transcribe is
reached through the pass-through route rather than the Add Model form, so
it joins the frozen unlisted set the Add Model drift test tracks.
Python resolves the Vertex project and location as call params, then the
litellm.vertex_project / litellm.vertex_location globals, then env, and
Azure AD token refresh from litellm.enable_azure_ad_token_refresh alone.
Native OCR skipped the globals, so a config.yaml litellm_settings value
silently fell through to the credential's project and us-central1, and a
managed identity setup without an API key failed. The bridge now reads
them through a provider_defaults settings group into OcrSettings, and
VertexConfig / AzureAuthInputs slot them in at Python's precedence.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
load_config used to return a local general_settings dict that it had
normalized in place, turning the configured role_permissions entries into
RoleBasedPermissions objects. It now returns the SettingsStore, which never
saw that write, so JWT auth received raw dicts and every request failed with
"'dict' object has no attribute 'role'" whenever role_permissions was set.
Convert the entries in the consumer instead, with a TypeAdapter, so the value
is parsed wherever it comes from. load_config keeps validating at boot, so a
malformed entry still fails startup rather than the first request.
Python reads every provider credential fallback (MISTRAL_API_KEY,
AZURE_AI_API_KEY, AZURE_DOCUMENT_INTELLIGENCE_ENDPOINT, Azure AD and
Vertex env, ...) through get_secret_str, which consults the configured
key_management_system before os.environ. Native OCR read std::env
directly, so a key held only in the vault went missing and a stale env
copy silently won. OcrClient now carries an injected secret Lookup that
the connection exposes to providers and auth crates; the bridge backs it
with settings.secret -> get_secret_str, pure Rust keeps the process env.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A pre-signed batch/file request (Mistral, Bedrock) carries its auth header
inside the transformed request body, which pre_call logs verbatim into
raw_request_typed_dict and raw-request callbacks, leaking the provider key.
Mask the nested headers channel before handing the request to pre_call.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
litellm-http now builds the rustls config itself, so one route-neutral place covers roots, the client certificate, ALPN, ssl_ecdh_curve and ssl_security_level. A curve picks the single key exchange group. A cipher string restricts the TLS 1.2 suites it names, and entries rustls cannot express, such as @SECLEVEL=1, are logged once and skipped.
user_url_validation and user_url_allowed_hosts are applied by the media fetcher. Document downloads honor the environment proxy whenever provider calls do, keeping the per-hop address check, and stay on the pinned resolver when no proxy applies.
AIOHTTP_SO_KEEPALIVE, AIOHTTP_TCP_KEEPIDLE, AIOHTTP_TCP_KEEPINTVL, AIOHTTP_TCP_KEEPCNT and AIOHTTP_KEEPALIVE_TIMEOUT map onto the client. A client= argument and a live SSLContext are ignored
Python becomes a thin SDK interface over Rust, so a live Python HTTP client has no effect on either route. This puts the Python OCR path back to what main does
The proxy attaches its shared aiohttp session to every request as shared_session, so declining on it sent every proxy OCR call to Python, which never uses that session for OCR. aclient_session is a litellm global and never a call argument, so that check could not match. The proxy-shaped lifecycle test now asserts the call was served by Rust
