Skip to content

fix(azure): forward realtime_protocol from config and relax api_version check for GA path#22369

Merged
krrishdholakia merged 43 commits intoBerriAI:litellm_oss_staging_02_28_2026from
giulio-leone:fix/issue-22127-azure-realtime-protocol
Feb 28, 2026
Merged

fix(azure): forward realtime_protocol from config and relax api_version check for GA path#22369
krrishdholakia merged 43 commits intoBerriAI:litellm_oss_staging_02_28_2026from
giulio-leone:fix/issue-22127-azure-realtime-protocol

Conversation

@giulio-leone
Copy link

Summary

Fixes #22127

Azure's GA v1 realtime API path (/openai/v1/realtime) is unusable when configured via config.yaml because:

  1. realtime_protocol set in litellm_params is not reliably forwarded to the handler
  2. api_version is unconditionally required, but the GA/v1 path doesn't use it

Changes

1. realtime_api/main.py — Add fallback chain for realtime_protocol

realtime_protocol = (
    kwargs.get("realtime_protocol")
    or litellm_params.get("realtime_protocol")     # NEW: from config.yaml extra fields
    or os.environ.get("LITELLM_AZURE_REALTIME_PROTOCOL")  # NEW: env var fallback
    or "beta"
)

2. llms/azure/realtime/handler.py — Conditional api_version check

# Before: unconditionally required (blocks GA path)
if api_version is None:
    raise ValueError("api_version is required for Azure OpenAI calls")

# After: only required for beta protocol
if api_version is None and realtime_protocol not in ("GA", "V1"):
    raise ValueError("api_version is required for Azure OpenAI calls")

3. Case-insensitive protocol matching

ga, GA, v1, V1 all work correctly now.

4. Type signature fix

_construct_url(api_version: str)_construct_url(api_version: Optional[str]) since GA path doesn't need it.

Tests

17 tests pass (7 existing + 10 new):

  • test_async_realtime_ga_without_api_version — GA path works without api_version
  • test_async_realtime_beta_without_api_version_raises — beta still requires api_version
  • test_realtime_protocol_env_var_fallback — env var fallback works
  • test_realtime_protocol_from_litellm_params — config extra field works
  • test_construct_url_case_insensitive_protocol[ga/Ga/gA/V1/v1/GA] — case insensitivity

Backwards Compatible

  • Default behavior unchanged (realtime_protocol="beta")
  • Existing deployments unaffected

ZeroClover and others added 30 commits February 25, 2026 01:29
…tion

Add headers parameter to image_generation() and aimage_generation() methods
in OpenAI provider, and pass headers from images/main.py to ensure custom
headers like cf-aig-authorization are properly forwarded to the OpenAI API.
Aligns behavior with completion() method and Azure provider implementation.
Verify that extra_headers are correctly forwarded to OpenAI's
images.generate() in both sync and async paths, and that they
are absent when not provided.
When a gunicorn worker exits (e.g. from max_requests recycling), its
per-process prometheus .db files remain on disk. For gauges using
livesum/liveall mode, this means the dead worker's last-known values
persist as if the process were still alive. Wire gunicorn's child_exit
hook to call mark_process_dead() so live-tracking gauges accurately
reflect only running workers.
…cleanup

Add Prometheus child_exit cleanup for gunicorn workers
…ng, and LLM Gateway (#21130)

* docs: update AssemblyAI docs with Universal-3 Pro, Speech Understanding, and LLM Gateway provider config

* feat: add AssemblyAI LLM Gateway as OpenAI-compatible provider
…th_info

Tests were mocking the old method name `filter_server_ids_by_ip` but production
code at server.py:774 calls `filter_server_ids_by_ip_with_info` which returns
a (server_ids, blocked_count) tuple. The unmocked method on AsyncMock returned
a coroutine, causing "cannot unpack non-iterable coroutine object" errors.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…ion behavior

Tests were asserting no response.create/conversation.item.create sent to
backend when guardrail blocks, but the implementation intentionally sends
these to have the LLM voice the guardrail violation message to the user.

Updated assertions to verify the correct guardrail flow:
- response.cancel is sent to stop any in-progress response
- conversation.item.create with violation message is injected
- response.create is sent to voice the violation
- original blocked content is NOT forwarded

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The revert in 8565c70 removed the parallel_tool_calls handling from
map_openai_params, and the subsequent fix d0445e1 only re-added the
transform_request consumption but forgot to re-add the map_openai_params
producer that sets _parallel_tool_use_config. This meant parallel_tool_calls
was silently ignored for all Bedrock models.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Commit 99c62ca removed "azure" from _RESPONSES_API_PROVIDERS,
routing Azure models through litellm.completion instead of
litellm.responses. The test was not updated to match, causing it
to assert against the wrong mock.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…22319)

* feat: add in_flight_requests metric to /health/backlog + prometheus

* refactor: clean class with static methods, add tests, fix sentinel pattern

* docs: add in_flight_requests to prometheus metrics and latency troubleshooting
PR #22271 added the LiteLLM_ClaudeCodePluginTable model to
schema.prisma but did not include a corresponding migration file,
causing test_aaaasschema_migration_check to fail.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Addresses Greptile review feedback.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…sertions

fix(test): update realtime guardrail test assertions for voice violation behavior
…e-test

fix(test): update Azure pass-through test after Responses API routing change
fix(db): add missing migration for LiteLLM_ClaudeCodePluginTable
…s-map-params

fix(bedrock): restore parallel_tool_calls mapping in map_openai_params
…ove_key to prevent unawaited coroutine warnings

Fixes #22128
…e agents (#22329)

* fix: enforce RBAC on agent endpoints — block non-admin create/update/delete

- Add /v1/agents/{agent_id} to agent_routes so internal users can
  access GET-by-ID (previously returned 403 due to missing route pattern)
- Add _check_agent_management_permission() guard to POST, PUT, PATCH,
  DELETE agent endpoints — only PROXY_ADMIN may mutate agents
- Add user_api_key_dict param to delete_agent so the role check works
- Add comprehensive unit tests for RBAC enforcement across all roles

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix: mock prisma_client in internal user get-agent-by-id test

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* feat(ui): hide agent create/delete controls for non-admin users

Match MCP servers pattern: wrap '+ Add New Agent' button in
isAdmin conditional so internal users see a read-only agents view.
Delete buttons in card and table were already gated.
Update empty-state copy for non-admin users.
Add 7 Vitest tests covering role-based visibility.

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>
…d-name

fix(mcp): update test mocks for renamed filter_server_ids_by_ip_with_info
* fix: Add PROXY_ADMIN role to system user for key rotation

The key rotation worker was failing with 'You are not authorized to regenerate this key'
when rotating team keys. This was because the system user created by
get_litellm_internal_jobs_user_api_key_auth() was missing the user_role field.

Without user_role=PROXY_ADMIN, the system user couldn't bypass team permission checks
in can_team_member_execute_key_management_endpoint(), causing authorization failures
for team key rotation.

This fix adds user_role=LitellmUserRoles.PROXY_ADMIN to the system user, allowing
it to bypass team permission checks and successfully rotate keys for all teams.

* test: Add unit test for system user PROXY_ADMIN role

- Verify internal jobs system user has PROXY_ADMIN role
- Critical for key rotation to bypass team permission checks
- Regression test for PR #21896
…2239)

* fix: populate user_id and user_info for admin users in /user/info endpoint

Fixes #22179

When admin users call /user/info without a user_id parameter, the endpoint
was returning null for both user_id and user_info fields. This broke
budgeting tooling that relies on /user/info to look up current budget and spend.

Changes:
- Modified _get_user_info_for_proxy_admin() to accept user_api_key_dict parameter
- Added logic to fetch admin's own user info from database
- Updated function to return admin's user_id and user_info instead of null
- Updated unit test to verify admin user_id is populated

The fix ensures admin users get their own user information just like regular users.

* test: make mock get_data signature match real method

- Updated MockPrismaClientDB.get_data() to accept all parameters that the real method accepts
- Makes mock more robust against future refactors
- Added datetime and Union imports
- Mock now returns None when user_id is not provided
…ed-coroutine

fix(caching): store task references in LLMClientCache._remove_key
fix(image_generation): propagate extra_headers to Upstream
…onses and chat completions (#22291)

* fixed dynamic auth for /responses with mcp

* fixed greptile concern
Fixes #18381: When using both tools and response_format with Bedrock
Converse API, LiteLLM internally adds json_tool_call to handle structured
output. Bedrock may return both this internal tool AND real user-defined
tools, breaking consumers like OpenAI Agents SDK.

Changes:
- Non-streaming: Added _filter_json_mode_tools() to handle 3 scenarios:
  only json_tool_call (convert to content), mixed (filter it out), or
  no json_tool_call (pass through)
- Streaming: Added json_mode tracking to AWSEventStreamDecoder to suppress
  json_tool_call chunks and convert to text content
- Fixed optional_params.pop() mutation issue

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Addresses review comment from greptile-apps:
#21107 (review)

Changes:
- Added `_unwrap_bedrock_properties()` helper method to eliminate code duplication
- Replaced two identical JSON unwrapping blocks (lines 1592-1601 and 1612-1620)
  with calls to the new helper method
- Improves maintainability - single source of truth for Bedrock properties unwrapping logic

The helper method:
- Parses JSON string
- Checks for single "properties" key structure
- Unwraps and returns the properties value
- Returns original string if unwrapping not needed or parsing fails

No functional changes - pure refactoring.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fixed MyPy errors where BedrockConverseConfig was used instead of
AmazonConverseConfig in the _unwrap_bedrock_properties() calls.

Errors:
- Line 1619: BedrockConverseConfig -> AmazonConverseConfig
- Line 1631: BedrockConverseConfig -> AmazonConverseConfig

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
…th support

Fixes #21941

The generated result filenames from _save_confusion_results contained
parentheses, dots, and full yaml filenames, producing paths that exceed
the Windows 260-char MAX_PATH limit. Rework the safe_label logic to
produce short {topic}_{method_abbrev} filenames (e.g. insults_cf.json)
while preserving the full label inside the JSON content.

Rename existing tracked result files to match the new naming convention.
…r/guardrail_benchmarks/test_eval.py

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
fix: shorten guardrail benchmark result filenames for Windows long path support
rasmi and others added 13 commits February 27, 2026 19:33
* fix(mcp): default available_on_public_internet to true

MCPs were defaulting to private (available_on_public_internet=false) which
was a breaking change. This reverts the default to public (true) across:
- Pydantic models (AddMCPServerRequest, UpdateMCPServerRequest, LiteLLM_MCPServerTable)
- Prisma schema @default
- mcp_server_manager.py YAML config + DB loading fallbacks
- UI form initialValue and setFieldValue defaults

* fix(ui): add forceRender to Collapse.Panel so toggle defaults render correctly

Ant Design's Collapse.Panel lazy-renders children by default. Without
forceRender, the Form.Item for 'Available on Public Internet' isn't
mounted when the useEffect fires form.setFieldValue, causing the Switch
to visually show OFF even though the intended default is true.

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix(mcp): update remaining schema copies and MCPServer type default to true

Missed in previous commit per Greptile review:
- schema.prisma (root)
- litellm-proxy-extras/litellm_proxy_extras/schema.prisma
- litellm/types/mcp_server/mcp_server_manager.py MCPServer class

* ui(mcp): reframe network access as 'Internal network only' restriction

Replace scary 'Available on Public Internet' toggle with 'Internal network only'
opt-in restriction. Toggle OFF (default) = all networks allowed. Toggle ON =
restricted to internal network only. Auth is always required either way.

- MCPPermissionManagement: new label/tooltip/description, invert display via
  getValueProps/getValueFromEvent so underlying available_on_public_internet
  value is unchanged
- mcp_server_view: 'Public' → 'All networks', 'Internal' → 'Internal only' (orange)
- mcp_server_columns: same badge updates

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>
…all-scoped

fix(bedrock): filter internal json_tool_call when mixed with real tools
…or hints (#22336)

* fix(jwt): support OIDC discovery URLs, handle roles array, improve error hints

Three fixes for Azure AD JWT auth:

1. OIDC discovery URL support - JWT_PUBLIC_KEY_URL can now be set to
   .well-known/openid-configuration endpoints. The proxy fetches the
   discovery doc, extracts jwks_uri, and caches it.

2. Handle roles claim as array - when team_id_jwt_field points to a list
   (e.g. AAD's "roles": ["team1"]), auto-unwrap the first element instead
   of crashing with 'unhashable type: list'.

3. Better error hint for dot-notation indexing - when team_id_jwt_field is
   set to "roles.0" or "roles[0]", the 401 error now explains to use
   "roles" instead and that LiteLLM auto-unwraps lists.

* Add integration demo script for JWT auth fixes (OIDC discovery, array roles, dot-notation hints)

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add demo_servers.py for manual JWT auth testing with mock JWKS/OIDC endpoints

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add demo screenshots for PR comment

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add integration test results with screenshots for PR review

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* address greptile review feedback (greploop iteration 1)

- fix: add HTTP status code check in _resolve_jwks_url before parsing JSON
- fix: remove misleading bracket-notation hint from debug log (get_nested_value does not support it)

* Update tests/test_litellm/proxy/auth/test_handle_jwt.py

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* remove demo scripts and assets

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
)

* perf: raise aiohttp connection pool limits (300→1000, 50/host→500)

* perf: skip model_copy() on every chunk — only copy usage-bearing chunks

* perf: replace list+join O(n²) with str+= O(n) in async_data_generator

* perf: cache model-level guardrail lookup per request, not per chunk
Add 88 tests across 9 test files for the CostTrackingSettings component directory:
- provider_display_helpers.test.ts: 9 tests for helper functions
- how_it_works.test.tsx: 9 tests for discount calculator component
- add_provider_form.test.tsx: 7 tests for provider form validation
- add_margin_form.test.tsx: 9 tests for margin form with type toggle
- provider_discount_table.test.tsx: 12 tests for table editing and interactions
- provider_margin_table.test.tsx: 13 tests for margin table with sorting
- use_discount_config.test.ts: 11 tests for discount hook logic
- use_margin_config.test.ts: 12 tests for margin hook logic
- cost_tracking_settings.test.tsx: 15 tests for main component and role-based rendering

All tests passing. Coverage includes form validation, user interactions, API calls, state management, and conditional rendering.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
[Test] UI - CostTrackingSettings: Add comprehensive Vitest coverage
Add filtering capabilities to /key/list endpoint for project_id and access_group_id parameters. Both filters work globally across all visibility rules and stack with existing sort/pagination params. Added comprehensive unit tests for the new filters.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
- Add ProjectDetailsPage with header, details card, spend/budget progress,
  model spend bar chart, keys placeholder, and team info card
- Refactor CreateProjectModal into base form pattern (ProjectBaseForm)
  shared between Create and Edit flows
- Add EditProjectModal with pre-filled form data from backend
- Add useProjectDetails and useUpdateProject hooks
- Add duplicate key validation for model limits and metadata
- Wire project ID click in table to navigate to detail view
- Move pagination inline with search bar

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
[Feature] Key list endpoint: Add project_id and access_group_id filters
…ateProjectModal.tsx

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
[Feature] UI - Projects: Add Project Details Page
…on check for GA path

The realtime_protocol parameter set in config.yaml litellm_params was
not reliably reaching the Azure realtime handler. Add fallback chain:
kwargs → litellm_params → LITELLM_AZURE_REALTIME_PROTOCOL env var → beta.

Also relax the api_version validation to only require it for the beta
protocol path, since the GA/v1 path does not use api_version in the URL.

Make protocol matching case-insensitive so 'ga', 'GA', 'v1', 'V1' all
work consistently. Fix _construct_url type signature to accept Optional
api_version.

Fixes #22127

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
@vercel
Copy link

vercel bot commented Feb 28, 2026

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
litellm Ready Ready Preview, Comment Feb 28, 2026 7:16am

Request Review

@CLAassistant
Copy link

CLAassistant commented Feb 28, 2026

CLA assistant check
All committers have signed the CLA.

@greptile-apps
Copy link
Contributor

greptile-apps bot commented Feb 28, 2026

Greptile Summary

This PR fixes Azure's GA (v1) realtime API path by:

  • Adding a fallback chain for realtime_protocol (kwargslitellm_params → env var → "beta") so it can be set via config.yaml
  • Relaxing the api_version validation in the handler to allow None when using GA/V1 protocol
  • Making protocol matching case-insensitive and updating the type signature accordingly

The handler changes are clean and correct. Two areas need attention:

  • The api_version default in main.py (line 131) always resolves to "2024-10-01-preview", so the handler's new GA relaxation is unreachable from the main proxy flow — consider gating the default to beta-only
  • Two new tests (test_realtime_protocol_env_var_fallback, test_realtime_protocol_from_litellm_params) duplicate the fallback logic inline instead of actually calling _arealtime, so they don't verify the real function's behavior

Confidence Score: 4/5

  • This PR is safe to merge — default behavior is unchanged and GA path is correctly handled at the handler level.
  • Core handler changes are well-implemented with case-insensitive matching and proper type signature updates. The api_version default chain in main.py makes the handler relaxation unreachable from the main flow (not a bug but limits the fix's value). Two tests don't exercise the actual function under test. No risk to existing functionality.
  • litellm/realtime_api/main.py (api_version default interacts with GA path), tests/test_litellm/llms/azure/realtime/test_azure_realtime_handler.py (two tests don't call _arealtime)

Important Files Changed

Filename Overview
litellm/llms/azure/realtime/handler.py Relaxes api_version validation for GA/V1 protocol, adds case-insensitive protocol matching, and updates _construct_url signature to Optional[str]. Changes are clean and correct.
litellm/realtime_api/main.py Adds litellm_params and env var fallback for realtime_protocol. The api_version default chain always provides a non-None value, making the handler's GA relaxation unreachable from this path.
tests/test_litellm/llms/azure/realtime/test_azure_realtime_handler.py Good handler-level tests (GA URL, case-insensitivity, beta validation). However, test_realtime_protocol_env_var_fallback and test_realtime_protocol_from_litellm_params don't actually call _arealtime — they duplicate its logic inline, so they don't verify the real function.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["_arealtime() in main.py"] --> B{Resolve realtime_protocol}
    B --> C["kwargs.get('realtime_protocol')"]
    C -->|None| D["litellm_params.get('realtime_protocol')"]
    D -->|None| E["os.environ LITELLM_AZURE_REALTIME_PROTOCOL"]
    E -->|None| F["Default: 'beta'"]
    C -->|found| G[protocol resolved]
    D -->|found| G
    E -->|found| G
    F --> G

    G --> H{Resolve api_version}
    H --> I["api_version or litellm_params.api_version or '2024-10-01-preview'"]
    I --> J["api_version always non-None"]

    J --> K["handler.async_realtime()"]
    K --> L{api_version is None AND protocol not GA/V1?}
    L -->|Yes| M["Raise ValueError"]
    L -->|No| N["_construct_url()"]
    N --> O{protocol.upper in GA, V1?}
    O -->|Yes| P["wss://.../openai/v1/realtime?model=..."]
    O -->|No| Q["wss://.../openai/realtime?api-version=...&deployment=..."]
Loading

Last reviewed commit: 994998d

Copy link
Contributor

@greptile-apps greptile-apps bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

3 files reviewed, 2 comments

Edit Code Review Agent Settings | Greptile

Comment on lines +308 to +327
@pytest.mark.asyncio
async def test_realtime_protocol_env_var_fallback():
"""
Test that LITELLM_AZURE_REALTIME_PROTOCOL env var is used as fallback.
Fixes #22127: no way to set realtime_protocol from config.
"""
from litellm.realtime_api.main import _arealtime
from litellm.types.router import GenericLiteLLMParams

with patch.dict(os.environ, {"LITELLM_AZURE_REALTIME_PROTOCOL": "v1"}):
# Create a GenericLiteLLMParams without realtime_protocol
litellm_params = GenericLiteLLMParams()
# The env var should be picked up as fallback
realtime_protocol = (
{}.get("realtime_protocol")
or litellm_params.get("realtime_protocol")
or os.environ.get("LITELLM_AZURE_REALTIME_PROTOCOL")
or "beta"
)
assert realtime_protocol == "v1"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tests duplicate logic instead of testing _arealtime

test_realtime_protocol_env_var_fallback imports _arealtime but never calls it. It manually re-implements the fallback chain inline (lines 321-326), so it only proves the Python or chain works — not that _arealtime actually reads LITELLM_AZURE_REALTIME_PROTOCOL. If someone later removes the env var fallback from _arealtime, this test would still pass.

The same issue applies to test_realtime_protocol_from_litellm_params (line 330): it only tests that GenericLiteLLMParams stores extra fields, not that _arealtime consumes realtime_protocol from those params.

Consider making these tests call _arealtime (mocking the Azure handler) and asserting that the handler receives the correct realtime_protocol value.

@greptile-apps
Copy link
Contributor

greptile-apps bot commented Feb 28, 2026

Additional Comments (1)

litellm/realtime_api/main.py
api_version default makes GA relaxation unreachable here

The api_version fallback chain on lines 128-132 always resolves to a non-None value (defaulting to "2024-10-01-preview"). This means the relaxed api_version is None check added in handler.py:89 will never be triggered from this code path — the handler always receives a valid api_version, even when the GA path doesn't use it.

This isn't a runtime bug (the GA URL is built correctly because _construct_url ignores api_version for GA), but it is misleading: users on the GA path who deliberately omit api_version from their config will silently get "2024-10-01-preview" injected. Consider gating the default to only apply for the beta protocol:

        realtime_protocol = (
            kwargs.get("realtime_protocol")
            or litellm_params.get("realtime_protocol")
            or os.environ.get("LITELLM_AZURE_REALTIME_PROTOCOL")
            or "beta"
        )

        api_version = (
            api_version
            or litellm_params.api_version
            or (None if realtime_protocol.upper() in ("GA", "V1") else "2024-10-01-preview")
        )

@krrishdholakia krrishdholakia changed the base branch from main to litellm_oss_staging_02_28_2026 February 28, 2026 08:03
@krrishdholakia krrishdholakia merged commit ff60394 into BerriAI:litellm_oss_staging_02_28_2026 Feb 28, 2026
25 of 30 checks passed
Sameerlite pushed a commit that referenced this pull request Mar 2, 2026
…on check for GA path (#22369)

* fix(image_generation): propagate extra_headers to OpenAI image generation

Add headers parameter to image_generation() and aimage_generation() methods
in OpenAI provider, and pass headers from images/main.py to ensure custom
headers like cf-aig-authorization are properly forwarded to the OpenAI API.
Aligns behavior with completion() method and Azure provider implementation.

* test(image_generation): add tests for extra_headers propagation

Verify that extra_headers are correctly forwarded to OpenAI's
images.generate() in both sync and async paths, and that they
are absent when not provided.

* Add Prometheus child_exit cleanup for gunicorn workers

When a gunicorn worker exits (e.g. from max_requests recycling), its
per-process prometheus .db files remain on disk. For gauges using
livesum/liveall mode, this means the dead worker's last-known values
persist as if the process were still alive. Wire gunicorn's child_exit
hook to call mark_process_dead() so live-tracking gauges accurately
reflect only running workers.

* docs: update AssemblyAI docs with Universal-3 Pro, Speech Understanding, and LLM Gateway (#21130)

* docs: update AssemblyAI docs with Universal-3 Pro, Speech Understanding, and LLM Gateway provider config

* feat: add AssemblyAI LLM Gateway as OpenAI-compatible provider

* fix(mcp): update test mocks to use renamed filter_server_ids_by_ip_with_info

Tests were mocking the old method name `filter_server_ids_by_ip` but production
code at server.py:774 calls `filter_server_ids_by_ip_with_info` which returns
a (server_ids, blocked_count) tuple. The unmocked method on AsyncMock returned
a coroutine, causing "cannot unpack non-iterable coroutine object" errors.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(test): update realtime guardrail test assertions for voice violation behavior

Tests were asserting no response.create/conversation.item.create sent to
backend when guardrail blocks, but the implementation intentionally sends
these to have the LLM voice the guardrail violation message to the user.

Updated assertions to verify the correct guardrail flow:
- response.cancel is sent to stop any in-progress response
- conversation.item.create with violation message is injected
- response.create is sent to voice the violation
- original blocked content is NOT forwarded

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(bedrock): restore parallel_tool_calls mapping in map_openai_params

The revert in 8565c70 removed the parallel_tool_calls handling from
map_openai_params, and the subsequent fix d0445e1 only re-added the
transform_request consumption but forgot to re-add the map_openai_params
producer that sets _parallel_tool_use_config. This meant parallel_tool_calls
was silently ignored for all Bedrock models.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(test): update Azure pass-through test to mock litellm.completion

Commit 99c62ca removed "azure" from _RESPONSES_API_PROVIDERS,
routing Azure models through litellm.completion instead of
litellm.responses. The test was not updated to match, causing it
to assert against the wrong mock.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add in_flight_requests metric to /health/backlog + prometheus (#22319)

* feat: add in_flight_requests metric to /health/backlog + prometheus

* refactor: clean class with static methods, add tests, fix sentinel pattern

* docs: add in_flight_requests to prometheus metrics and latency troubleshooting

* fix(db): add missing migration for LiteLLM_ClaudeCodePluginTable

PR #22271 added the LiteLLM_ClaudeCodePluginTable model to
schema.prisma but did not include a corresponding migration file,
causing test_aaaasschema_migration_check to fail.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update stale docstring to match guardrail voicing behavior

Addresses Greptile review feedback.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix(caching): store background task references in LLMClientCache._remove_key to prevent unawaited coroutine warnings

Fixes #22128

* [Feat] Agent RBAC Permission Fix - Ensure Internal Users cannot create agents (#22329)

* fix: enforce RBAC on agent endpoints — block non-admin create/update/delete

- Add /v1/agents/{agent_id} to agent_routes so internal users can
  access GET-by-ID (previously returned 403 due to missing route pattern)
- Add _check_agent_management_permission() guard to POST, PUT, PATCH,
  DELETE agent endpoints — only PROXY_ADMIN may mutate agents
- Add user_api_key_dict param to delete_agent so the role check works
- Add comprehensive unit tests for RBAC enforcement across all roles

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix: mock prisma_client in internal user get-agent-by-id test

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* feat(ui): hide agent create/delete controls for non-admin users

Match MCP servers pattern: wrap '+ Add New Agent' button in
isAdmin conditional so internal users see a read-only agents view.
Delete buttons in card and table were already gated.
Update empty-state copy for non-admin users.
Add 7 Vitest tests covering role-based visibility.

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix: Add PROXY_ADMIN role to system user for key rotation (#21896)

* fix: Add PROXY_ADMIN role to system user for key rotation

The key rotation worker was failing with 'You are not authorized to regenerate this key'
when rotating team keys. This was because the system user created by
get_litellm_internal_jobs_user_api_key_auth() was missing the user_role field.

Without user_role=PROXY_ADMIN, the system user couldn't bypass team permission checks
in can_team_member_execute_key_management_endpoint(), causing authorization failures
for team key rotation.

This fix adds user_role=LitellmUserRoles.PROXY_ADMIN to the system user, allowing
it to bypass team permission checks and successfully rotate keys for all teams.

* test: Add unit test for system user PROXY_ADMIN role

- Verify internal jobs system user has PROXY_ADMIN role
- Critical for key rotation to bypass team permission checks
- Regression test for PR #21896

* fix: populate user_id and user_info for admin users in /user/info (#22239)

* fix: populate user_id and user_info for admin users in /user/info endpoint

Fixes #22179

When admin users call /user/info without a user_id parameter, the endpoint
was returning null for both user_id and user_info fields. This broke
budgeting tooling that relies on /user/info to look up current budget and spend.

Changes:
- Modified _get_user_info_for_proxy_admin() to accept user_api_key_dict parameter
- Added logic to fetch admin's own user info from database
- Updated function to return admin's user_id and user_info instead of null
- Updated unit test to verify admin user_id is populated

The fix ensures admin users get their own user information just like regular users.

* test: make mock get_data signature match real method

- Updated MockPrismaClientDB.get_data() to accept all parameters that the real method accepts
- Makes mock more robust against future refactors
- Added datetime and Union imports
- Mock now returns None when user_id is not provided

* [Fix] Pass MCP auth headers from request into tool fetch for /v1/responses and chat completions (#22291)

* fixed dynamic auth for /responses with mcp

* fixed greptile concern

* fix(bedrock): filter internal json_tool_call when mixed with real tools

Fixes #18381: When using both tools and response_format with Bedrock
Converse API, LiteLLM internally adds json_tool_call to handle structured
output. Bedrock may return both this internal tool AND real user-defined
tools, breaking consumers like OpenAI Agents SDK.

Changes:
- Non-streaming: Added _filter_json_mode_tools() to handle 3 scenarios:
  only json_tool_call (convert to content), mixed (filter it out), or
  no json_tool_call (pass through)
- Streaming: Added json_mode tracking to AWSEventStreamDecoder to suppress
  json_tool_call chunks and convert to text content
- Fixed optional_params.pop() mutation issue

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* refactor: extract duplicated JSON unwrapping into helper method

Addresses review comment from greptile-apps:
#21107 (review)

Changes:
- Added `_unwrap_bedrock_properties()` helper method to eliminate code duplication
- Replaced two identical JSON unwrapping blocks (lines 1592-1601 and 1612-1620)
  with calls to the new helper method
- Improves maintainability - single source of truth for Bedrock properties unwrapping logic

The helper method:
- Parses JSON string
- Checks for single "properties" key structure
- Unwraps and returns the properties value
- Returns original string if unwrapping not needed or parsing fails

No functional changes - pure refactoring.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix: use correct class name AmazonConverseConfig in helper method calls

Fixed MyPy errors where BedrockConverseConfig was used instead of
AmazonConverseConfig in the _unwrap_bedrock_properties() calls.

Errors:
- Line 1619: BedrockConverseConfig -> AmazonConverseConfig
- Line 1631: BedrockConverseConfig -> AmazonConverseConfig

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

* fix: shorten guardrail benchmark result filenames for Windows long path support

Fixes #21941

The generated result filenames from _save_confusion_results contained
parentheses, dots, and full yaml filenames, producing paths that exceed
the Windows 260-char MAX_PATH limit. Rework the safe_label logic to
produce short {topic}_{method_abbrev} filenames (e.g. insults_cf.json)
while preserving the full label inside the JSON content.

Rename existing tracked result files to match the new naming convention.

* Update litellm/proxy/guardrails/guardrail_hooks/litellm_content_filter/guardrail_benchmarks/test_eval.py

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* Remove Apache 2 license from SKILL.md (#22322)

* fix(mcp): default available_on_public_internet to true (#22331)

* fix(mcp): default available_on_public_internet to true

MCPs were defaulting to private (available_on_public_internet=false) which
was a breaking change. This reverts the default to public (true) across:
- Pydantic models (AddMCPServerRequest, UpdateMCPServerRequest, LiteLLM_MCPServerTable)
- Prisma schema @default
- mcp_server_manager.py YAML config + DB loading fallbacks
- UI form initialValue and setFieldValue defaults

* fix(ui): add forceRender to Collapse.Panel so toggle defaults render correctly

Ant Design's Collapse.Panel lazy-renders children by default. Without
forceRender, the Form.Item for 'Available on Public Internet' isn't
mounted when the useEffect fires form.setFieldValue, causing the Switch
to visually show OFF even though the intended default is true.

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix(mcp): update remaining schema copies and MCPServer type default to true

Missed in previous commit per Greptile review:
- schema.prisma (root)
- litellm-proxy-extras/litellm_proxy_extras/schema.prisma
- litellm/types/mcp_server/mcp_server_manager.py MCPServer class

* ui(mcp): reframe network access as 'Internal network only' restriction

Replace scary 'Available on Public Internet' toggle with 'Internal network only'
opt-in restriction. Toggle OFF (default) = all networks allowed. Toggle ON =
restricted to internal network only. Auth is always required either way.

- MCPPermissionManagement: new label/tooltip/description, invert display via
  getValueProps/getValueFromEvent so underlying available_on_public_internet
  value is unchanged
- mcp_server_view: 'Public' → 'All networks', 'Internal' → 'Internal only' (orange)
- mcp_server_columns: same badge updates

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* fix(jwt): OIDC discovery URLs, roles array handling, dot-notation error hints (#22336)

* fix(jwt): support OIDC discovery URLs, handle roles array, improve error hints

Three fixes for Azure AD JWT auth:

1. OIDC discovery URL support - JWT_PUBLIC_KEY_URL can now be set to
   .well-known/openid-configuration endpoints. The proxy fetches the
   discovery doc, extracts jwks_uri, and caches it.

2. Handle roles claim as array - when team_id_jwt_field points to a list
   (e.g. AAD's "roles": ["team1"]), auto-unwrap the first element instead
   of crashing with 'unhashable type: list'.

3. Better error hint for dot-notation indexing - when team_id_jwt_field is
   set to "roles.0" or "roles[0]", the 401 error now explains to use
   "roles" instead and that LiteLLM auto-unwraps lists.

* Add integration demo script for JWT auth fixes (OIDC discovery, array roles, dot-notation hints)

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add demo_servers.py for manual JWT auth testing with mock JWKS/OIDC endpoints

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add demo screenshots for PR comment

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* Add integration test results with screenshots for PR review

Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>

* address greptile review feedback (greploop iteration 1)

- fix: add HTTP status code check in _resolve_jwks_url before parsing JSON
- fix: remove misleading bracket-notation hint from debug log (get_nested_value does not support it)

* Update tests/test_litellm/proxy/auth/test_handle_jwt.py

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* remove demo scripts and assets

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* perf: streaming latency improvements — 4 targeted hot-path fixes (#22346)

* perf: raise aiohttp connection pool limits (300→1000, 50/host→500)

* perf: skip model_copy() on every chunk — only copy usage-bearing chunks

* perf: replace list+join O(n²) with str+= O(n) in async_data_generator

* perf: cache model-level guardrail lookup per request, not per chunk

* test: add comprehensive Vitest coverage for CostTrackingSettings

Add 88 tests across 9 test files for the CostTrackingSettings component directory:
- provider_display_helpers.test.ts: 9 tests for helper functions
- how_it_works.test.tsx: 9 tests for discount calculator component
- add_provider_form.test.tsx: 7 tests for provider form validation
- add_margin_form.test.tsx: 9 tests for margin form with type toggle
- provider_discount_table.test.tsx: 12 tests for table editing and interactions
- provider_margin_table.test.tsx: 13 tests for margin table with sorting
- use_discount_config.test.ts: 11 tests for discount hook logic
- use_margin_config.test.ts: 12 tests for margin hook logic
- cost_tracking_settings.test.tsx: 15 tests for main component and role-based rendering

All tests passing. Coverage includes form validation, user interactions, API calls, state management, and conditional rendering.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>

* [Feature] Key list endpoint: Add project_id and access_group_id filters

Add filtering capabilities to /key/list endpoint for project_id and access_group_id parameters. Both filters work globally across all visibility rules and stack with existing sort/pagination params. Added comprehensive unit tests for the new filters.

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>

* [Feature] UI - Projects: Add Project Details page with Edit modal

- Add ProjectDetailsPage with header, details card, spend/budget progress,
  model spend bar chart, keys placeholder, and team info card
- Refactor CreateProjectModal into base form pattern (ProjectBaseForm)
  shared between Create and Edit flows
- Add EditProjectModal with pre-filled form data from backend
- Add useProjectDetails and useUpdateProject hooks
- Add duplicate key validation for model limits and metadata
- Wire project ID click in table to navigate to detail view
- Move pagination inline with search bar

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Update ui/litellm-dashboard/src/components/Projects/ProjectModals/CreateProjectModal.tsx

Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>

* fix(azure): forward realtime_protocol from config and relax api_version check for GA path

The realtime_protocol parameter set in config.yaml litellm_params was
not reliably reaching the Azure realtime handler. Add fallback chain:
kwargs → litellm_params → LITELLM_AZURE_REALTIME_PROTOCOL env var → beta.

Also relax the api_version validation to only require it for the beta
protocol path, since the GA/v1 path does not use api_version in the URL.

Make protocol matching case-insensitive so 'ga', 'GA', 'v1', 'V1' all
work consistently. Fix _construct_url type signature to accept Optional
api_version.

Fixes #22127

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

---------

Co-authored-by: Zero Clover <zero@root.me>
Co-authored-by: Ryan Crabbe <rcrabbe@berkeley.edu>
Co-authored-by: ryan-crabbe <128659760+ryan-crabbe@users.noreply.github.com>
Co-authored-by: Dylan Duan <dylan.duan@assemblyai.com>
Co-authored-by: Julio Quinteros Pro <jquinter@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: Shivaang <shivaang.05@gmail.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Ishaan Jaff <ishaan-jaff@users.noreply.github.com>
Co-authored-by: milan-berri <milan@berri.ai>
Co-authored-by: Shivam Rawat <161387515+shivamrawat1@users.noreply.github.com>
Co-authored-by: Brian Caswell <bcaswell@microsoft.com>
Co-authored-by: Brian Caswell <bcaswell@gmail.com>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
Co-authored-by: rasmi <rrelasmar@gmail.com>
Co-authored-by: yuneng-jiang <yuneng.jiang@gmail.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Azure Realtime API — realtime_protocol config param not forwarded, GA v1 path unusable