* Warn when budget lookup fails; cache won't populate
- Add _log_budget_lookup_failure helper in auth_checks.py
- Log at WARNING in get_user_object, get_team_object, get_key_object
when DB lookups fail (schema mismatch, etc.)
- Add schema migration hint for prisma/db errors
- Add dry-run test for _log_budget_lookup_failure
* fix: skip budget lookup failure log for expected user-not-found case
Avoid logging 'cache will not be populated' when the user simply doesn't
exist - not caching is correct behavior in that case. Only log for
unexpected errors (schema, DB, etc.) where the message is meaningful.
Cache request.url.path once instead of accessing it 6 times (2 in
assistants check + 4 in LITELLM_METADATA_ROUTES loop). Reduces
per-call time from 47µs to 20µs (-58%).
Also inline the assistants API check to avoid function call overhead.
Adds 8 tests for _get_metadata_variable_name covering all return paths.
Create fresh mock objects within the test instead of reusing mocks
from setUp that have side_effect configured. The setUp's side_effect
on mock_langfuse_client.trace can interfere with return_value settings
when tests try to reset and reconfigure mocks.
Using dedicated mock objects for this test avoids state pollution
from setUp's side_effect configuration and makes the test more
deterministic in parallel execution environments.
Set cached tokenizer config directly and mock both sync and async
tokenizer functions to avoid race conditions when running with
parallel test execution (-n 16).
The issue was that parallel tests could populate the
litellm.known_tokenizer_config cache between clearing it and
when the code checked it. This caused the sync code path to be
used instead of the async path, bypassing the mocked async functions.
Fix:
1. Set cache directly instead of clearing it
2. Also mock sync versions _get_tokenizer_config and _get_chat_template_file
This ensures the test is deterministic regardless of test execution order.
Unify follow-up fixes for Opus 4.6 pricing and routing metadata into
a single changeset.
Set long-context-capable Opus 4.6 entries to 1M input tokens where
>200K pricing is defined, align alias and dated capability metadata,
and add Bedrock Converse v1 IDs with and without :0 suffixes.
Keep regional endpoint pricing at a 10% premium over global entries
and mirror all cost-map changes in the backup file used for local
loading and offline fallback behavior.
Extend Opus 4.6 regression tests to verify metadata parity, Bedrock
regional pricing parity across :0 and non-:0 IDs, and converse model
registration in constants and runtime model sets.
Add Claude Opus 4.6 entries for Anthropic, Bedrock Converse, and Vertex AI.
Align pricing and capability metadata with Anthropic docs, including
long-context rates, above-200k prompt-caching rates, prefill removal,
and tool-use system prompt token counts.
Register the Bedrock Converse model ID in constants and add targeted
tests to validate model map values and converse registration.
* _translate_openai_content_to_anthropic
* test_response_format_consistency
* test fixes unit tests
* test fix
* fix: use dict access for anthropic content blocks in tests (#20447)
The translate_openai_response_to_anthropic method returns dicts, not objects.
Changed .type/.text/.thinking attribute access to dict ['key'] access.
---------
Co-authored-by: shin-bot-litellm <shin-bot-litellm@berri.ai>
* fix(a2a): use text/event-stream SSE format for message/stream endpoint
The A2A gateway's streaming response was using application/x-ndjson
Content-Type and raw NDJSON body format. The A2A protocol spec requires
text/event-stream with SSE framing (data: ...\n\n).
The official a2a-sdk client validates the Content-Type header and raises
SSEError when it doesn't contain text/event-stream.
Changes:
- Changed media_type from application/x-ndjson to text/event-stream
- Updated response body to use SSE framing (data: prefix + \n\n suffix)
- Added tests validating Content-Type and SSE body format
Fixes#20278
* Potential fix for code scanning alert no. 4045: Information exposure through an exception
Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com>
* test(a2a): add manual SSE format validation script
Adds a manual test script that:
1. Starts a real A2A agent on port 10001
2. Starts LiteLLM proxy with the agent registered
3. Makes a streaming request to the proxy's A2A gateway
4. Validates Content-Type header is text/event-stream
5. Validates body uses SSE framing (data: ...\n\n)
Run: python tests/a2a_manual/test_a2a_sse_manual.py
---------
Co-authored-by: shin-bot-litellm <shin-bot-litellm@users.noreply.github.com>
Co-authored-by: Ishaan Jaff <ishaanjaffer0324@gmail.com>
Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com>
The unified guardrail's streaming iterator hook processes every Nth
chunk (sampling_rate, default 5). On each sampled chunk it calls
process_output_streaming_response, which combines all accumulated text
into the first chunk and clears all subsequent chunks to "".
The hook then yielded `processed_items[-1]` — the last item, whose
content had been cleared to "". This permanently lost every Nth
chunk's content, causing random missing words/tokens in the client
output (observed in Roo Code, Open WebUI, etc.).
Fix: deep-copy the current chunk before guardrail processing runs,
then yield the original (unmodified) chunk. The guardrail validation
still executes and can block if it detects a problem, but the stream
content is preserved.
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
