PR #32272 added a separate models_mgmt suite for the /model/new, /model/update,
and /model/delete lifecycle; the management suite that merged later covers the
sibling key/team/user/organization routes, and two suites for one route family
is one too many. Fold the four model tests in as a second test file, merge the
model methods into ManagementClient (find_deployment, update_model_tpm, a
strict delete_model_strict matching the delete_key_strict naming, and the raw
create_model_status), and port the typed pieces to models.py: DeploymentParams
mirroring LiteLLM_Params so /model/info assertions can reach any configured
knob, tpm on LiteLLMParamsBody, and ModelUpdateParams/ModelUpdateBody.
All 12 suite tests verified live against the split stage deployment
UI session tokens carry the virtual team_id litellm-dashboard (UI_TEAM_ID),
which is never persisted. The MCP team-permission helpers passed it to
get_team_object anyway, so every dashboard MCP listing raised a 404 per
lookup that was swallowed into per-server 'Failed to get allowed tools for
server' warnings (plus the sibling 'allowed MCP servers for team' and 'MCP
access groups for team' warnings) and wasted DB queries. The 404 also
escaped past the key-level permission handling in
get_allowed_tools_for_server, dropping key tool restrictions for such
sessions.
Short-circuit the virtual team before the DB lookup in the three helpers,
mirroring the existing UI_TEAM_ID handling in agent_permission_handler.
Also reject /team/new with the reserved team_id, since a real row would
bind its budget and permissions to every UI session
* fix(main): stop per-request custom pricing from clobbering shared model_cost pricing
A request routed through a wildcard deployment with explicit zero pricing
(e.g. openai/* with input_cost_per_token: 0) registered that pricing on the
shared {provider}/{model} key in litellm.model_cost, so sibling deployments
relying on built-in pricing logged $0 until process restart (LIT-3991).
Request-time registration in completion()/embedding() now mirrors the
router-startup isolation: router-originated requests register full pricing
under the deployment's unique model id only, while the shared backend key
receives the entry with custom pricing fields stripped. Direct SDK calls
without a router deployment id keep the legacy shared-key registration.
The stripping logic is shared via
CustomPricingLiteLLMParams.strip_custom_pricing_fields and reused by
Router._create_deployment and Router.add_deployment.
* test: update legacy tests that asserted per-request pricing leaking into shared model_cost
test_router_fallbacks_with_custom_model_costs asserted the shared
claude-sonnet-4-5-20250929 entry ends up with the deployment's 30/60
pricing, which is exactly the cross-deployment leak this PR removes; it
now asserts the shared key keeps the built-in pricing, matching the
test's stated goal.
test_cost_calc.py::test_run computed streaming cost via
completion_cost(response), which only matched the non-stream cost while
the shared gpt-3.5-turbo entry was poisoned with the per-request
2/token pricing; it now passes the request's custom pricing explicitly
via custom_cost_per_token.
* refactor(mcp): read oauth2_flow verbatim from DB rows; inference stays config-only plus a logged backstop
With every DB write site stamping oauth2_flow (#32283, #32288) and the startup
backfill healing legacy null rows, the DB build no longer needs to re-derive the
flow from field shape. build_mcp_server_from_table now reads the column verbatim
via _explicit_oauth2_flow: unknown or null values resolve to None, which
needs_user_oauth_token already treats as interactive, so an unstamped row degrades
to the safe default instead of guessing M2M from a shape that a DCR-registered
interactive server shares whenever discovery is down
Field-shape inference survives in exactly two places. config.yaml-loaded servers
keep it at load time: they are rebuilt from the config on every boot, so there is
no row to backfill and load-time resolution is their write-time stamp. And the
request-time backstop in _get_allowed_mcp_servers keeps a not-yet-backfilled M2M
row blocking caller Authorization forwarding (the P1 property); it now logs a
warning whenever it actually fires, which is the fire-rate signal for deleting it
once deployments have booted past the backfill
Regression tests pin that the DB build does not infer M2M from the credential
shape and reads an explicit column value verbatim
Fourth step of the oauth2_flow persistence sequence, stacked on the backfill
* feat(mcp): deprecation warning when config-level M2M is inferred rather than declared
A config.yaml oauth2 server whose credential shape decides client_credentials without
an explicit oauth2_flow now logs a warning at load pointing the admin at the explicit
declaration. First rung of the deprecation ladder: the docs make oauth2_flow the
recommended path, the warning surfaces configs still relying on inference, and a
future breaking release can turn it into a config validation error, at which point
config-level shape inference dies entirely. Interactive omissions stay silent since
the default matches inference there and nothing load-bearing is being guessed
* feat(mcp)!: require explicit oauth2_flow for config-defined oauth2 servers
A config.yaml server with auth_type oauth2 must now declare its flow; the load
raises a config validation error naming both values and what each means:
oauth2_flow: client_credentials for machine-to-machine (the proxy mints a shared
token at token_url using client_id/client_secret) or
oauth2_flow: authorization_code for interactive (per-user tokens via browser
sign-in, including delegate_auth_to_upstream)
This replaces the load-time shape inference for config servers entirely. The
credential shape is genuinely ambiguous (a DCR-registered interactive server
carries client creds + token_url with no authorization_url, identical to M2M),
so the config asserts the answer instead of the proxy guessing it. With this,
field-shape inference survives in exactly one place: the request-time security
backstop, which is telemetry-gated for deletion
BREAKING CHANGE: config-defined oauth2 MCP servers without oauth2_flow fail
proxy startup with the error above. Add the one line to the server block; the
error text says exactly which value to pick
* test(mcp): pin the verbatim read for authorization_code alongside client_credentials
Raised by review on the PR
* fix(mcp): fail closed on the anonymous delegate gate for unstamped M2M-shaped servers
Reading oauth2_flow verbatim (this PR) changed has_client_credentials from True to
False for a legacy null-flow row that still carries the M2M credential shape. That
value is what the anonymous upstream-delegate gate checks before skipping LiteLLM
auth entirely, so an M2M-shaped delegate server that was never stamped would newly
pass the gate: an unauthenticated caller could get it selected and then list/read
upstream data using the client credentials the request-time backstop re-infers,
running as LiteLLM's service account. This reopens the hole the gate's existing
'never delegate for M2M' guard was written to close
The gate now resolves the flow (column first, shape fallback) instead of reading the
bare column, mirroring the request-time backstop in _get_allowed_mcp_servers: both
fail closed on the ambiguous M2M shape and are removed together once no null rows
remain. A pure-PKCE delegate server (no stored credentials) resolves to a non-M2M
flow and keeps its bypass, so the common delegate case is unaffected
Tests: an unstamped M2M-shaped delegate server is denied the bypass (mutation-checked
against the bare-column regression), and a pure-PKCE delegate server still bypasses
Raised by review on the PR
* fix(mcp): centralize the request-time oauth2_flow backstop across every security site
Reading oauth2_flow verbatim made has_client_credentials unreliable for legacy null
rows, and the backstop that compensates was applied at only one reader. Review found
three more consequences of that per-site approach:
- the anonymous-delegate allowlist in get_allowed_mcp_servers read the bare column, so
an unstamped M2M-shape delegate server was surfaced to anonymous callers (High)
- call_mcp_tool resolved allowed ids into MCPServer objects without the backstop, so a
null-flow M2M-shape row kept has_client_credentials false on tool execution during a
backfill gap, though the listing path was covered (High)
- the request-time warning claimed the startup backfill would stamp the row next boot,
but the backfill deliberately leaves the ambiguous M2M shape unstamped (Low)
Rather than patch each site, introduce two helpers on MCPServerManager that are the
single choke point for request-time resolution: effective_oauth2_flow(server) for the
enum/boolean decisions (allowlist filter, anonymous-delegate gate) and
resolve_oauth2_flow_for_request(server) for the egress object copy (listing and tool
call). Both fail closed on the M2M shape and leave stamped rows and pure-PKCE rows
untouched. The gate now shares effective_oauth2_flow instead of its inline resolution,
and the corrected warning lives once inside resolve_oauth2_flow_for_request, so deleting
the whole transitional layer later is a single-site change.
Tests: helper unit coverage (stamped verbatim, null M2M-shape resolves, pure-PKCE stays
None, stamped/pure-PKCE return the same object, corrected warning text), the anonymous
allowlist excludes an unstamped M2M-shape delegate server, and the call path resolves
the flow like the listing path. The two security-integration tests are mutation-checked
against the bare-column regression.
Raised by review on the PR
* fix(spend): bound the logs-tab pagination count to stop full-window scans
The spend-logs UI list endpoint (/spend/logs/ui, /spend/logs/v2) computed an
exact pagination total over the whole selected time window on every load. That
was a standalone SELECT COUNT(*) FROM (SELECT request_id FROM LiteLLM_SpendLogs
WHERE ...) which, on a multi-TB SpendLogs table, scans a huge number of rows and
spikes Aurora ACU. The later LIT-4027 change folded it into COUNT(*) OVER (), but
a window count still drains every matching row before the LIMIT applies, so the
full-window scan remained.
Compute the total with a bounded SELECT COUNT(*) FROM (SELECT 1 ... LIMIT $cap+1)
that probes at most cap+1 rows, and drop the window count from the page query so
the page query is a plain indexed ORDER BY ... LIMIT. When more than the cap
match, report the cap and set total_is_capped so the UI renders "<cap>+". The
bounded subquery terminates early rather than aggregating across all tablets, so
it stays safe on sharded engines like YugabyteDB too.
Resolves LIT-4119
* test(spend): make zero-total mock reflect real COUNT(*), add capped-total tooltip
Address Greptile review on #31825:
- the empty-result test now returns [{"total_count": 0}] for the bounded
count query (real COUNT(*) always returns one row) instead of [], so the
zero-total path exercises the normal branch rather than the defensive guard
- the logs toolbar shows a tooltip explaining the cap when total_is_capped is
set, so a disabled Next button at the cap boundary reads as intentional
Add an opt-in mode so a key that exceeds its own max_budget is throttled to a
globally configured percentage of its TPM/RPM instead of being blocked entirely.
A new litellm_settings global, budget_exceeded_throttle_percentage, sets the
fraction (e.g. 0.1 = 10%). A per-key throttle_on_budget_exceeded flag (stored in
key metadata via the existing management-endpoint metadata routing) opts the key
in. When both are set and the key is over budget, the budget check records the
percentage on a request-scoped budget_throttle_pct instead of raising, and the
rate limiter scales the key's configured TPM/RPM by it. Keys without the flag
keep hard-blocking; team/user/org budgets are unaffected.
The throttle is recomputed from the key's original limits on every request and
the decision is cleared before the auth object is cached, so it never compounds
across requests. Both the budget read-time check and the budget reservation path
honor the opt-in, and both the v3 and legacy rate limiters apply the scaling.
Enabling throttle_on_budget_exceeded is proxy-admin only. It converts an
admin-imposed hard budget block into a soft throttle that keeps spending past
max_budget, so a non-admin must not be able to self-opt-in and bypass their own
spend cap. Both /key/generate and /key/update reject a non-admin setting it to
true (update only gates the transition to enabled, so a non-admin can still edit
other fields and turn the flag off). This matches the feature being wholly
proxy-admin operated: the global percentage is admin-only too.
A key that opts in but has no TPM or RPM limit has nothing to scale, so it stays
hard-blocked rather than serving unlimited requests past its budget (fail-safe).
The global budget_exceeded_throttle_percentage is configurable from the admin UI
(Settings -> General Settings), persisted through litellm_settings so it survives
a restart, not only from config.yaml.
Resolves LIT-3894. Scope for LIT-3893.
* fix(mcp): apply semantic filter to expanded litellm_proxy tools and show filtered-out count
* fix(mcp): guard expansion-path filtering on enabled flag and isolate metadata emission
* fix(mcp): resolve tool name prefix via known server prefixes, not string match
When an MCP server's alias differs from its server_name, tool names are
listed with the alias prefix but _execute_tool_calls compared that prefix
against the server_name stored in tool_server_map. The mismatch silently
skipped prefix stripping, forwarding the fully-prefixed tool name upstream
and causing "Unknown tool" failures. Resolve the actual MCPServer object
and strip using its known prefix forms (alias, server_name, server_id)
instead.
* fix(mcp): preserve tool overrides and scope REST tool listing
Return saved tool display/description overrides from the server table API
so the edit UI reloads them, resolve display names before prefix stripping
on tool calls, and honor mcp_server_name and toolset_name filters on the
REST tools list endpoint.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(mcp): inject BYOK credentials on Playground OpenAPI tool calls
Playground and Responses API route MCP execution through call_tool, which
skipped BYOK lookup and never set the OpenAPI auth ContextVar, so upstream
calls went out unauthenticated despite a stored user credential.
Co-authored-by: Cursor <cursoragent@cursor.com>
* test(mcp): cover alias-mismatch prefix stripping and display-name reverse mapping
Regression tests for _execute_tool_calls: an MCP server whose alias differs
from its server_name must still have its tool-name prefix stripped correctly,
and a tool called by its configured display name must resolve back to the
original tool name before dispatch.
* fix(mcp): validate tool display names against Bedrock's tool-name pattern
A display name replaces the tool name sent to the LLM provider, so a value
with spaces or other special characters saves successfully but fails every
subsequent Bedrock tool call. Validate tool_name_to_display_name server-side
(create/update payload) against Bedrock's [a-zA-Z0-9_-]+ constraint, and add
matching inline validation plus a save-blocking guard in the Admin UI's
create and edit MCP server forms.
* style(mcp): fix ruff/prettier formatting on CI
No logic changes; satisfies the format checks flagged on PR #32320.
* fix(mcp): fix CI failures on PR - complexity budget and stale test mock
Extract toolset-scope resolution and query-param normalization out of
list_tool_rest_api into helpers to bring it back under the C901 complexity
budget (was 18, now within the 15 threshold).
Add the missing get_mcp_server_by_name stub to the streaming iterator test's
mock manager; the alias-fallback resolution added for tool-name-prefix
stripping calls it unconditionally when _get_mcp_server_from_tool_name misses.
* test(mcp): cover BYOK OpenAPI auth-header helpers to close codecov patch gap
_format_byok_openapi_auth_header, _openapi_forwarded_extra_headers, and
_resolve_byok_mcp_auth_header were only exercised indirectly via a mocked
call_tool test, leaving their branches (auth-type formatting, header
forwarding/stripping, missing-credential 401) uncovered.
* fix(mcp): resolve BYOK auth before queuing the during-hook task
_resolve_byok_mcp_auth_header can raise a 401 when no credential is stored.
Resolving it after during_hook_task was already queued meant a hook's
side effects (audit logging, rate-limit bookkeeping) could run and record
success for a tool call that then fails on the missing credential.
* fix: correct mcp alias routing regressions
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(proxy): wire general_settings SSRF allowlist to litellm globals
general_settings.user_url_allowed_hosts was documented in SSRF errors but
never applied at startup, so internal MCP/OpenAPI URLs stayed blocked.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(proxy): regenerate dashboard types and satisfy ruff UP006 budget
Use list[str] in ConfigGeneralSettings and run gen:api so schema.d.ts
matches the new SSRF general_settings fields.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix: normalize ssrf general settings
* fix: clear ssrf allowlists from null settings
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
Once the active Prisma client is in the disconnected state, every DB call
raises ClientNotConnectedError. The reconnect machinery was supposed to
recover from this, but _get_engine_pid() inspected the broken client via
prisma's _engine property, which re-raises that same error, so
recreate_prisma_client failed before it could build a replacement client
and the proxy looped on failed reconnects forever (issue #28322 showed
1486+ consecutive failures over 30 days with zero recoveries)
Guard both _get_engine_pid implementations with is_connected() so a
disconnected client reads as "no engine" (pid 0) and the recreate path
proceeds to construct and connect a fresh client
* fix(caching): pass only metadata to valkey semantic async embedding
ValkeySemanticCache async get/set passed **kwargs into _get_async_embedding,
which raised TypeError on cache_key and other fields and silently skipped
all cache writes. Match redis-semantic by forwarding metadata only.
Co-authored-by: Cursor <cursoragent@cursor.com>
* test(caching): add async_get_cache embedding call regression test
Mirror the async_set_cache spy test so async_get_cache passing **kwargs
into _get_async_embedding is caught by a real signature, not AsyncMock.
Co-authored-by: Cursor <cursoragent@cursor.com>
---------
Co-authored-by: Shivam Rawat <shivamrawat@Shivams-MacBook-Pro.local>
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(batches): price anthropic passthrough message batches correctly in batch cost job
Anthropic message batches created via the /anthropic passthrough were never
cost tracked. The CheckBatchCost job fetched batch results from the Files API
(POST /v1/files/msgbatch_.../content), which Anthropic rejects with "File id
must have file_ prefix"; the error response was silently wrapped as file
content, parsed as zero successful rows, logged as a $0 aretrieve_batch spend
row, and the job was marked batch_processed=true so the $0 was permanent.
Route msgbatch_ file ids to GET /v1/messages/batches/{id}/results in the
anthropic files transformation, raise on HTTP error status in
retrieve_file_content instead of returning the error body as content, parse
Anthropic's results JSONL shape (result.type == "succeeded",
result.message.usage with cache creation/read tokens) in batch_utils, price
cache creation tokens at cache_creation_input_token_cost in the batch cost
fallback (50% batch discount preserved for base input, cache reads, cache
writes, and output), and leave the managed object row unprocessed when cost
tracking fails so a later poll retries instead of permanently recording $0.
* fix(batches): carry cache token details into aggregated anthropic batch usage
The x-litellm-semantic-filter-tools response header was sliced mid-name at
MAX_MCP_SEMANTIC_FILTER_TOOLS_HEADER_LENGTH with a trailing "...", so the
admin UI test panel rendered the last selected tool name chopped. Truncate
the CSV at a tool name boundary instead so the header only ever carries
complete names, and note in the test panel how many selected tools did not
fit in the header
Streaming pass-through for native Anthropic /v1/messages and the /v1/responses
streaming iterator never set logging_obj.completion_start_time, so
_success_handler_helper_fn fell back to completion_start_time = end_time.
Downstream TTFT consumers (Prometheus, OTEL, Langfuse, Admin UI, spend logs
completionStartTime) then reported time-to-first-token equal to total request
duration.
Stamp completion_start_time on the first chunk in PassThroughStreamingHandler.
chunk_processor and BaseResponsesAPIStreamingIterator._process_chunk, mirroring
CustomStreamWrapper for /chat/completions.
Resolves LIT-4185
Co-authored-by: yucheng <yucheng@yuchengs-MBP.localdomain>
* test(e2e): add management suite covering key/team/user/org lifecycle and route permissions
* test(e2e): decouple the enforcement-flip assertion from upstream health
Polling for a 200 on the newly-allowed model required it to be a routable,
healthy upstream, which is not the contract under test; poll until the
key_model_access_denied 403 lifts instead, excluding 401 so a revoked key
cannot read as success. Also document that the delete test's deferred teardown
firing on an already-deleted key is deliberate: cleanup must survive the test
failing before the in-body delete, and the repeat delete is a warn-free no-op
(the proxy answers 404 No keys found)
* test(e2e): inline the management suite's model and tpm literals
* test(e2e): drop the models_mgmt suite line from the folder list
* test(e2e): write the tpm limit as a plain integer literal
* feat(mcp): startup backfill stamping oauth2_flow on legacy null rows
Rows created before the write-side stamps carry a null oauth2_flow and rely on
read-time field-shape inference, which cannot tell a DCR-registered interactive
server (client creds + token_url, no persisted authorization_url) from an M2M
server unless endpoint discovery succeeds first; on a transient discovery failure
those servers flip to client_credentials for that registry load
The backfill classifies each null oauth2 row once, at rest, ordered by signal
strength: per-user token rows (only the interactive flow mints them, so this is
definitive and catches the DCR-trap cohort), then a persisted authorization_url,
then a persisted registration_url (DCR implies interactive; this covers
registered-but-never-signed-in rows), then the M2M credential shape mirroring the
legacy inference, else the interactive default that matches how
needs_user_oauth_token treats a null flow. Every stamp is logged with the rule
that fired and written with updated_by=oauth2_flow_backfill for auditability
Runs in _init_mcp_servers_in_db before the registry load so the first build of
the boot classifies from the column, is isolated so a failure cannot block server
loading, and is idempotent: a healed fleet exits after one indexed query. This
unblocks deleting the read-time inference for DB rows in the follow-up
Third step of the oauth2_flow persistence sequence, after #32283 and #32288
* fix(mcp): backfill leaves the ambiguous M2M shape unstamped instead of guessing client_credentials
The credential shape (client_id + client_secret + token_url, no interactive signal) is
shared by real M2M servers and DCR-registered interactive servers nobody has signed
into: the DCR persist writes creds and token_url but not authorization_url or
registration_url. Stamping client_credentials from that shape permanently mislabeled
the interactive cohort, and once explicit the value is authoritative, so per-user
traffic would run on the proxy's stored client credential with no discovery rescue
and no backstop (it only guards null rows)
The backfill now stamps only what it can prove. Interactive signals keep stamping
authorization_code; the ambiguous shape is left null with an actionable warning naming
the server and the fix (set oauth2_flow via the dashboard or PUT /v1/mcp/server). A
true M2M row keeps working per-request through the security backstop while the warning
nags; an interactive row keeps its Authorize button (null renders interactive), and one
completed sign-in creates the per-user token that stamps it authorization_code at the
next boot. Mirrors the config-level rule: M2M is asserted by a human, never guessed
Raised by review on the PR
* perf(mcp): batch the backfill stamps into one update_many per flow value
The per-row update loop issued one DB round-trip per legacy row at startup; rows
sharing a stamped value now go out as a single update_many, so the DB cost is
constant in fleet size. Per-row logging keeps the rule that fired for each server
Raised by review on the PR
* fix(mcp): backfill stamps only rows still null at write time and counts only real OAuth token rows as sign-in proof
Two review findings. The batched update_many matched on server_id alone, so an
explicit oauth2_flow set between the backfill's read and its write (an admin PUT or
a sign-in's DCR stamp landing in the boot window) would be overwritten with the
inferred value; the where clause now also requires oauth2_flow to still be null, so
an explicit value can never be clobbered under any interleaving
And the per_user_tokens rule counted any LiteLLM_MCPUserCredentials row as proof of
an interactive sign-in, but that table doubles as BYOK storage for user-supplied API
keys; a BYOK-flavored row would have stamped an M2M-shaped server authorization_code.
The rule now counts only rows whose payload decodes as a type oauth2 token via the
existing _decode_oauth_payload discriminator, so bare keys, undecodable rows, and
stale leftovers from a BYOK-to-oauth2 auth switch prove nothing
Raised by review on the PR
Drive the real parallel_request_limiter through _pre_call_with_fallbacks for
the LIT-3890 customer scenario: a key-level model_tpm_limit raises
ProxyRateLimitError from the pre-call hook and the configured gateway fallback
serves the request instead of returning a 429. Unlike the existing tests, this
exercises the actual limiter rather than a hand-built error.
Also switch the new _pre_call_with_fallbacks return annotation to builtin
tuple to stay within the ruff UP006 strict-rule budget.
* fix(llm_http_handler): send dict transcription request data as a JSON body
httpx form-encodes dicts passed via data= and silently ignores json=, so the
generic audio transcription path never actually sent a JSON body. No provider
hit this before; JSON-body speech APIs need it.
* feat(vertex_ai): add Google Cloud Speech-to-Text Chirp 3 transcription support
Adds a VertexAIAudioTranscriptionConfig wired through ProviderConfigManager so
vertex_ai/chirp_3 works on /v1/audio/transcriptions (sync and async) via the
Speech-to-Text v2 recognize API. Auth reuses the standard Vertex credential
resolution (vertex_project/vertex_location/vertex_credentials or ADC); the
location defaults to the us multi-region since chirp_3 is only served from the
us and eu multi-regions, and non-global locations use the regional
<location>-speech.googleapis.com host. Maps language to languageCodes (auto
language detection by default), joins all result alternatives into the
transcript, and tracks cost from totalBilledDuration with a
vertex_ai/chirp_3 price entry at Google's published $0.016/min.
* fix(vertex_ai): map bare ISO-639-1 language codes to BCP-47 for Speech-to-Text
OpenAI clients send language codes like "en", which Google rejects with 400
("not supported by the model chirp_3 in the location us"); Speech-to-Text
wants region-qualified BCP-47 like "en-US". Adds a shared
normalize_transcription_language_to_bcp47 helper in audio_utils (NVIDIA Riva's
transcription config already hand-rolled the same table privately) that maps
common bare codes and passes region-qualified ones through, and applies it in
the Vertex transcription request. Also narrows the response JSON parse guard
to ValueError.
* fix(vertex_ai): drop zero output_cost_per_second so chirp_3 cost tracking works
cost_per_second prefers output_cost_per_second whenever it is not None, so the
0.0 in the chirp_3 entry priced every transcription at $0.00 instead of using
input_cost_per_second. Remove it from both cost maps and pin the behavior with
a regression test computing 18s of chirp_3 audio to ~$0.0048.
* fix(vertex_ai): validate client-controllable location to prevent SSRF in Speech-to-Text
get_complete_url interpolated vertex_location straight into the request host,
and vertex_location is client-controllable on the proxy (it flows from the
request body and is not on the request-body blocklist). An authenticated caller
could send vertex_location="attacker.example/" to point the host at their own
server, so the proxy would POST the audio plus its admin-minted Google bearer
token and x-goog-user-project header to the attacker, exfiltrating a
cloud-platform-scoped OAuth token minted from the admin's credentials.
Factor the location validation the rest of vertex_ai already applied in
get_vertex_base_url (^[a-z][a-z0-9-]*$ plus the global allowance) into a shared
validate_vertex_location helper in common_utils and call it from both the chat
host builder and the new speech host builder. Invalid locations now raise a 400
VertexAIError instead of building a host. Also reject vertex_project values that
carry URL-structural characters, since it lands in the URL path.
Regression tests assert on the parsed netloc so the security property is pinned:
valid locations always resolve to a *speech.googleapis.com host and injection
inputs are rejected.
* fix(vertex_ai): reject unsupported transcription response_format values instead of silently ignoring
* feat(mcp): persist oauth2_flow explicitly on create instead of inferring it at read time
The UI create payload never carried oauth2_flow, so every UI-created oauth2 server
persisted a null flow and relied on _resolve_oauth2_flow's field-shape inference at
registry build. That inference cannot tell a DCR-registered interactive server
(client creds + token_url, no persisted authorization_url) from an M2M server unless
endpoint discovery succeeds first, and the dashboard cannot reproduce it at all
because credentials are redacted in responses
The create form now persists the selected flow for oauth2 servers: authorization_code
for Interactive (PKCE), client_credentials for M2M. The REST create endpoints stamp an
omitted oauth2_flow server-side with the same discriminator the legacy inference uses,
run at write time where the payload carries plaintext credentials, so the decision is
made once with full information and stored. Applied to the admin create, the BYOM
submission, and the temporary session-server endpoints
The edit form derives its flow display from oauth2_flow instead of token_url presence
(token_url is present on authorization_code servers too, so it cannot distinguish M2M)
and deliberately never writes oauth2_flow: it has no flow selector, so a write from
edit could only erase an explicit value, including the authorization_code stamp the
DCR flow persists. Regression tests pin all of this down
Second step of persisting oauth2_flow at every write site so the legacy inference can
eventually be deleted; the backfill for existing null rows lands next
* refactor(mcp): name the create-time flow stamp for its fallback-only contract
stamp_omitted_oauth2_flow with a dedicated explicit-value early return and the shape
check renamed to has_m2m_shape, so the precedence (caller's oauth2_flow always wins,
inference only fills an omitted field) reads directly off the code
* fix(responses-bridge): custom tool round-trip and allowlist preservation for Codex CLI
Convert Responses API custom tools to Chat Completions function tools and map
function_call responses back to custom_tool_call output items so Codex CLI gets
the apply_patch round-trip it expects. Preserve and validate allowed_callers
during the custom->function conversion so the Anthropic adapter's caller
allowlist is not silently dropped, which would let a tool meant to be callable
only by another tool be invoked directly by the model. Use modern type
annotations (list/dict/set/X | None) throughout to keep the ruff strict budget
within its ratcheted ceilings.
* fix(responses-bridge): address review feedback on custom tool bridge
Type convert_custom_tool_to_function_tool against Mapping/ChatCompletionToolParam
and validate allowed_callers with a strict TypeAdapter so the two new cast()
calls that tripped the LIT006 ceiling are gone. Warn when dropping Responses-only
tool types (computer_use, image_generation, namespace, shell) instead of
discarding them silently. Return output items as Pydantic models instead of
model_dump()ing every item to a dict, matching the declared return type. Apply
the same None-safe metadata pattern to the request_data paths that still used
setdefault, and drop the unused build_custom_tool_call_item helper.
* fix(responses-bridge): recover custom tool input when arguments is empty
* fix(auth): extract custom tool names for allowlist enforcement on responses route
The Responses guardrail translation handler only extracted function and mcp
tool names, so a key or team restricted by metadata.allowed_tools could invoke
a disallowed tool by declaring it with type custom now that the bridge converts
custom tools into callable Chat Completions function tools. Extract custom tool
names through the same path so check_tools_allowlist rejects them.
* fix(responses-bridge): scope input payload recovery to custom_tool_call items
Recovering tool arguments from the input field on any falsy arguments value
made plain function_call input items with empty arguments and a stray input
key get rewritten into a {"content": ...} envelope, corrupting multi-turn
replay for normal function tools. Gate the recovery on the item type so it
only applies to custom_tool_call items, which are the ones that store their
payload in input.
* fix(responses-bridge): default missing function_call arguments to empty string
With input recovery scoped to custom_tool_call items, a plain function_call
input item without an arguments key left raw_arguments as None and the
downstream str() turned it into the literal string None. Coerce to an empty
string instead, matching the pre-bridge behavior.
---------
Co-authored-by: duanhongyi <duanhongyi@doopai.com>
* feat(jwt): fall back to DB team memberships when JWT has no team claims
* style(jwt): use PEP 585/604 annotations in DB team fallback to clear strict gate
* fix(jwt): preserve DB teams on no-claim sync, model-gate DB fallback, stop team-id leak
When fallback_to_db_teams is enabled and a JWT carries no team claims,
sync_user_role_and_teams previously computed teams_to_remove as every existing
DB membership and wiped the user out of all their teams on each request, which
also left the DB fallback nothing to resolve. Skip team removal in that case so
memberships survive and the fallback can attribute usage.
Apply the same per-team model-access check the claim-based path enforces when
selecting a DB fallback team, so a team's models restriction is no longer
bypassed; a team that cannot serve the requested model is skipped in favor of
one that can.
Drop the user's team-id list from the x-litellm-team-id membership 403 detail so
a valid-JWT caller can no longer enumerate team IDs.
* fix(jwt): load team membership on DB fallback; scope header check to provisional teams
The DB-team fallback resolved a team but never loaded its team membership
row, so per-team membership budget limits were silently skipped on that
path. _resolve_db_team_fallback now fetches the resolved team's membership
when a user_id is known and returns it, matching the claim-based path so
downstream LiteLLM_TeamMembership budget enforcement works there too.
The provisional x-litellm-team-id validation also fired on any non-None
team_id, including an RBAC role-derived one, which 403'd RBAC team flows
when the asserted team was not also a DB membership. It now runs only when
team_id actually came from the header (team_id == header_team_id).
* fix(jwt): surface DB-fallback membership lookup failures at warning level
A transient get_team_membership failure on the DB team fallback path is
recoverable: the team is still resolved and the request proceeds, just
without per-team membership budget enforcement for that request. Logging
that at debug hid a silent budget-enforcement gap from operators, so it now
logs at warning and states that enforcement was skipped. Behavior is
otherwise unchanged: the resolved team is returned with a None membership
rather than failing the request, covered by
test_resolve_db_team_fallback_survives_membership_lookup_error.
* fix(jwt-auth): tighten db-team fallback gating and passthrough enforcement
Resolves four issues in the fallback_to_db_teams path:
- _resolve_db_team_fallback now surfaces a model-access denial when memberships
exist but none can access the requested model, instead of always returning
the no-membership message
- auth_builder gates the fallback on real JWT team claims via
get_all_jwt_team_ids so a configured team_id_default does not silently route
claimless tokens to the default team
- A team selected only via _resolve_db_team_fallback is re-validated against
the team's allowed_passthrough_routes; the earlier gate ran while team_id
was still None
- sync_user_role_and_teams considers both plural and singular team claim
shapes when reconciling DB memberships so singular-only tokens
(Okta/Auth0 defaults) no longer leave stale teams behind
* fix(jwt): don't upsert a provisional x-litellm-team-id before membership check
When fallback_to_db_teams is on and the JWT carries no team claims, an
x-litellm-team-id header is accepted provisionally and only validated against
the user's DB memberships later in auth_builder. With team_id_upsert also
enabled, get_team_object ran the upsert on that unvalidated header team first,
so an attacker-supplied header could create an orphaned team row before the
403 membership check. Suppress the upsert whenever the team is provisional
(db_team_fallback), since a genuine membership team already exists and an
invalid one must not be created. Regression:
test_auth_builder_provisional_header_team_is_not_upserted.
* fix(jwt): pin RBAC-asserted team against db-team-fallback header override
When a JWT carries an RBAC team role but no group claims, auth_builder already
sets team_id from the RBAC object_id. db_team_fallback still evaluated true
there, so the provisional x-litellm-team-id path accepted a header team and
silently overrode the RBAC-asserted team with any team the caller belonged to.
Gate db_team_fallback on team_id being unset, and drive the header's provisional
acceptance off db_team_fallback rather than the raw flag, so an RBAC token plus
a non-claim header team is rejected with 403 instead of substituting the team.
Regression: test_auth_builder_header_cannot_override_rbac_team_under_db_fallback.
* fix(jwt): scope dual-claim membership sync to fallback_to_db_teams
The membership sync read both plural and singular JWT team claims via
get_all_jwt_team_ids unconditionally, which silently changed reconciliation
for every deployment using sync_user_role_and_teams, not just those opting
into fallback_to_db_teams: a singular-only IdP token that previously stripped
all DB teams would now be recognized. Gate the dual-claim read on
fallback_to_db_teams so flag-off deployments keep the upstream plural-only
behavior, honoring the PR's contract that existing deployments are unchanged.
Regression: test_sync_user_role_and_teams_singular_claim_only_recognized_under_flag.
* fix(jwt): drop user team IDs from db-fallback model-access 403 detail
The model-access-denied 403 in _resolve_db_team_fallback echoed the user's
full DB team-id list in its detail. It is only the caller's own memberships,
but it is inconsistent with the membership-validation 403 in the same feature
that was deliberately scrubbed of team IDs. Replace the enumerated list with a
generic "no team you are a member of has access" message. Regression extends
test_resolve_db_team_fallback_distinguishes_no_membership_vs_model_denied to
assert the team id is absent from the detail.
* fix(jwt): keep db-team fallback off for alias-only tokens
* test(jwt): cover alias-only token skipping db-team fallback
The autofix in ed21199 added a get_team_alias clause to the db_team_fallback
gate so an alias-only JWT (team_alias_jwt_field set, no team-id claims)
resolves its alias via find_and_validate_specific_team_id instead of being
mis-attributed to the user's first DB team, but it shipped without a
regression test. This drives auth_builder with an alias-only token whose
alias resolves to a different team than the user's DB membership and asserts
the result is the alias-resolved team; reverting the get_team_alias clause
flips the result to the DB-membership team, so the test fails without the fix
* fix(jwt): prefer alias resolution over team_id_default
When the JWT only carries an alias claim and the operator configures
team_id_default, JWTHandler.get_team_id silently substitutes the
default into find_and_validate_specific_team_id. That made the helper
return the default team without ever attempting alias resolution, so
spend and access attached to the default team even though the token
identified a different team via its alias. Use get_all_jwt_team_ids
(which ignores team_id_default) to detect when the resolved team_id is
only the default and clear it so alias resolution runs first; the
default remains the fallback when no alias claim is present.
* fix(jwt): enforce team_allowed_routes in db-team fallback resolution
The claim-based path runs allowed_routes_check when selecting a team, but
_resolve_db_team_fallback selected a team purely on model access, so a
DB-resolved team could reach routes excluded by team_allowed_routes with no
downstream backstop. This mirrors the claim path's route gate in the fallback,
exempting auth-enforced passthrough routes that are gated separately by
allowed_passthrough_routes at the call site
* fix(jwt): enforce team_allowed_routes on header-team db fallback path
The auto-pick DB-team fallback already gates against team_allowed_routes, but a claimless JWT presenting x-litellm-team-id under fallback_to_db_teams set team_id directly from the header and only re-validated DB membership afterwards, skipping the route gate. A caller could reach management/info routes that the JWT config narrowed for team-role callers by supplying the header even though the auto-pick path on the same route returns no team.
* refactor(jwt): narrow db-team fallback except clauses to actual failure types
* fix(jwt): collapse provisional header team lookup failure into membership denial
A caller holding a valid claimless JWT under fallback_to_db_teams could
distinguish nonexistent teams (404 from get_team_object) from existing
teams they do not belong to (membership 403) by varying x-litellm-team-id,
giving an authenticated team-id existence oracle. The provisional header
path now rewrites the lookup failure into the exact 403 the membership
check raises, while claim-backed header teams keep the upstream 404.
Also drop the unreachable falsy-team guard in _resolve_db_team_fallback
(get_team_object returns a team or raises, never None) and stop codecov
carryforward for three dead flags whose stale sessions were measured
against old file revisions and sank patch coverage with phantom
executable lines
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
ModelArmorGuardrail.async_pre_call_hook and async_moderation_hook hardcoded
their inner should_run_guardrail event type to pre_call / during_call. The
central dispatcher already remaps call_mcp_tool -> pre_mcp_call/during_mcp_call
and passes the outer gate, but Model Armor's redundant inner gate then rejected
MCP calls for a guardrail configured with mode pre_mcp_call/during_mcp_call, so
tool-call content was silently skipped.
Remap call_mcp_tool -> pre_mcp_call/during_mcp_call in both hooks, matching the
existing behavior of the noma and cisco guardrails. Adds regression tests
covering both hooks (scan runs on MCP calls, still skipped for chat traffic).
Generated with AI
Co-Authored-By: Claude Code
Co-authored-by: eugene-yao-zocdoc <eugene.yao@zocdoc.com>
* fix(mcp): forward short OAuth state upstream, keep session in a cookie
Some upstream authorization servers reject the OAuth authorize request with
"state parameter too long" because LiteLLM replaced the client's short state
with its own long encrypted session blob (base_url, original state, PKCE, client
redirect_uri) and sent that upstream as state.
Forward a short random handle as the upstream state instead, and carry the
encrypted session in a per-flow HttpOnly, SameSite=lax cookie bound to that
handle. The browser replays the cookie on /callback, so the session is recovered
without any server-side store and the client still gets its own original state
back. /callback falls back to decoding state directly when no cookie is present,
so flows in flight across a deploy keep working.
Resolves LIT-4197
* test(mcp): cover /callback error path cookie read and clear
The happy-path regression test already asserts the short-handle -> cookie round
trip. Add a focused test for the IdP-error branch of /callback: it must recover
the client's original state from the per-flow cookie (not the short handle),
propagate the error to the client's redirect_uri, and expire the one-time
cookie. Fails if the error path stops reading or clearing the cookie.
* fix(logging): classify allm_passthrough_route as async to prevent duplicate success callbacks
Async passthrough requests set kwargs["allm_passthrough_route"]=True but that
flag is never propagated into litellm_params, and _is_sync_litellm_request only
checks acompletion/aresponses/aembedding/aimage_generation/atranscription.
Every async passthrough is misclassified as sync, which trips the CustomLogger
sync branch in success_handler and fires log_success_event in addition to the
async worker's async_log_success_event, causing 2-3 duplicate LangSmith runs
per Bedrock passthrough request
Propagate allm_passthrough_route through get_litellm_params and teach the
classifier about it. /chat/completions and other non-passthrough paths are
untouched
* test(passthrough): assert allm_passthrough_route flag propagates end-to-end
Integration-level guard on top of the unit tests in test_litellm_logging.py:
verifies that when kwargs["allm_passthrough_route"]=True enters
llm_passthrough_route, the flag survives get_litellm_params(**kwargs), lands
in the logging object's litellm_params, and _is_sync_litellm_request reads
the request as async
---------
Co-authored-by: yucheng <yucheng@yuchengs-MBP.localdomain>
Only the gateway-managed interactive flow reaches this persist (the public /register
routes never pass persist_credentials), so the row it writes is authorization_code by
definition. It was not recorded, which left the row as client creds + token_url with
no persisted authorization_url and a null oauth2_flow: exactly the shape the legacy
M2M inference in _resolve_oauth2_flow matches. The row normally survives because
endpoint discovery backfills authorization_url in memory before the inference runs,
but on any transient discovery failure at registry build the server flips to
client_credentials for that load, routing per-user traffic to the M2M path
Stamping the flow at the write site makes the classification explicit and permanent,
so a DCR-registered interactive server no longer depends on discovery succeeding to
classify correctly. First step of persisting oauth2_flow at every write site so the
legacy inference can eventually be deleted
* fix(e2e): route model management to the control plane and restore Gateway.create_model
The split-transport routing table listed only /model/info as a control-plane
prefix, so /model/new and /model/delete were sent to the data-plane gateway,
which does not serve management routes and 404s them. Every suite that
registers deployments at runtime (llm_translation, batches, access_control)
failed on the split stage deployment because of this. Widen the prefix to
/model/ so all model-management routes reach the control plane while /models
stays on the data plane.
Separately, batch_client.py and several llm_translation tests call
gateway.create_model, but Gateway never had that method, so all 17 batch tests
errored at fixture setup with AttributeError. Add create_model/delete_model to
Gateway (with the optional mode that batches needs) and make EndpointsClient
delegate to it instead of carrying its own copy.
Regression tests cover both: the routing predicate for management vs LLM paths
and the Gateway model-management surface via a typed fake Transport. Both fail
on the previous code
* test(e2e): make the fake transport payload depend on response_type
The recording fake always answered with {"model_id": ...} even when the
caller asked for NoBody, which only validated because pydantic ignores extra
fields by default. Return an empty payload for response types that carry no
fields so a future extra="forbid" on NoBody cannot turn the delete test into
a ValidationError inside the fake
* test(e2e): probe the full spend read surface including schema-hidden routes
The curated spend-route list missed twelve read endpoints, most of them
include_in_schema=False and therefore invisible to the schema-discovery test:
/spend/logs/v2, /spend/logs/session/ui, /global/all_end_users,
/global/activity/exceptions/deployment, and the per-entity daily activity
family (user, user aggregated, team, organization, customer, end_user, tag).
Add them all, verified responsive against the live split stage deployment.
/end_user was missing from CONTROL_PLANE_PREFIXES, so /end_user/daily/activity
would have been routed to the data plane and 404ed like /model/new used to;
add the prefix and pin it plus the daily-activity routes in the transport
routing test.
/provider/budgets stays excluded with a documented reason: it returns 500
whenever router_settings.provider_budget_config is absent, so probing it on a
proxy without provider budget routing configured can never be green
* fix(security): hash Bearer-prefixed API keys in spend logs
The safety-net hash in get_logging_payload only checked for keys
starting with 'sk-', missing keys that arrived as 'Bearer sk-...'.
This caused plaintext API keys to be stored in SpendLogs for failed
requests while successful requests correctly stored SHA256 hashes.
Adds _hash_api_key_for_spend_log that strips the Bearer prefix
before hashing, applied to both the api_key column and the
metadata.user_api_key field in spend log payloads.
* fix: strip Bearer prefix from non-sk keys in spend log fallback path
---------
Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
CacheCodec.serialize dumped cached Pydantic models with model_dump(exclude_none=True), which drops any None-valued key, while deserialize does a strict model_validate. For a model with a required-but-nullable field (Optional[X] with no default), a None value is dropped on write and then fails model_validate on read with "Field required", so the entry can never be read back; that is a permanent cache miss, and in readers that rebuild the model from the raw cached dict an uncaught ValidationError that surfaces to the client as a 401
Removing exclude_none makes serialize and deserialize a lossless pair, so None fields are written as null and survive the round trip. LiteLLM_ManagedVectorStoresTable, the one cached model still carrying required-nullable fields and mis-caching on every read today, also gets the None defaults its peers already have
Round out the AWS auth-field coverage of _DB_LITELLM_PARAM_ENV_REF_KEYS
so every stringy aws_* field a deployment can pin in the DB resolves
os.environ/ refs at load time:
- aws_bedrock_project_id: Bedrock project/workspace association, banned
from request bodies via _BANNED_REQUEST_BODY_PARAMS
- aws_batch_role_arn: Bedrock batches role ARN (analog of aws_role_name)
- aws_workspace_id: Claude Platform workspace ID
Verified against three independent sources:
- BaseAWSLLM.aws_authentication_params (all 11)
- LiteLLM_Params-declared AWS fields (all 5)
- every aws_* string read from litellm_params/kwargs/optional_params
across litellm/ (all 14, excluding aws_bedrock_client which is a
boto3 client object, not a string, and aws_polly which is a provider
name)
The regression test now pins all 12 newly-allowlisted fields.
PR #30867 removed request-time os.environ/ expansion in
BaseAWSLLM.get_credentials to close LIT-3831. That relies on config-load
paths pre-resolving os.environ/ refs, but the DB-load path
(_resolve_db_litellm_param) only re-expands keys in
_DB_LITELLM_PARAM_ENV_REF_KEYS, which covered api_key,
aws_access_key_id, and aws_secret_access_key but not the other AWS auth
fields. A model stored in Postgres with e.g.
aws_role_name: os.environ/BEDROCK_ASSUME_ROLE_ARN
lands on the router with the literal string, get_credentials no longer
expands it, and STS returns
ValidationError: os.environ/BEDROCK_ASSUME_ROLE_ARN is invalid
Add the remaining AWS auth params to the allowlist so DB-sourced values
resolve at model-load time (trusted, server-side), matching the
YAML-config path. Team-scoped DB rows still get resolve_env_refs=False,
so the LIT-3831 defense-in-depth path is unchanged and request-body
injection is still blocked by _BANNED_REQUEST_BODY_PARAMS.
Regression tests pin every added field as an os.environ/ DB value and
assert it resolves on the router, plus a team-scoped pin that asserts
env refs remain literal.
Previously, every guardrail request forwarded the full conversation
history to CrowdStrike AIDR. In a multi-turn conversation this means
every prior message gets re-scanned on every new call, even though those
messages were already evaluated in earlier turns.
CrowdStrike AIDR internally has a conversation boundary optimization in
place for just this scenario (ref. <https://aidr-docs.crowdstrike.com/docs/aidr/apis#messages-array-optional---array-of-message-objects-containing-a-conversation-segment-with-the-ai-system>).
However, it is nevertheless wasteful to send so much data to the API
when only a subset of it will be processed. It also risks hitting the
documented 1 MiB request size limit.
So now we filter down to system messages plus either the messages after
the last assistant turn, or the last assistant message itself when that
is what is being guarded. We also preserve the original, full message
history within the guardrail in order to stitch back any
transformations.
Co-authored-by: Kenan Yildirim <kenan@kenany.me>
