- The Redis stall cell now reads the burst key's live counter and requires at least the full failure spend
- Unit tests drop the charging writer double, bare dicts, list mutation and wall-clock timestamps
- Partial-stream test models are frozen
* fix(http_handler): name the timeout phase and the limit that fired instead of 'after None seconds'
Every HTTP verb on both handlers raised litellm.Timeout with the per-call
timeout argument interpolated into the message, which is None whenever the
caller relied on the client default, and always said "Connection" even for a
read timeout. One helper now reads the limit httpx actually enforced from the
failed request and names the phase that fired
Fixes#14635
* fix(rust_bridge): name the timeout phase in the vault timeout message to match the httpx handler
* fix(http_handler): drop the unreachable timeout fallbacks httpx already rules out
* refactor(rust_bridge): drop a doc comment that restated the timeout helper
OAuthScopeResponse holds a tools/call response's http.response.start so a managed OAuth scope challenge can replace it with a 403. Once the call was allowed, the held start still waited for the next body message, which for a slow tool is the 15s SSE keep-alive ping. Send it as soon as the call is allowed instead.
Fixes#46659
* fix(router): stop concurrent deployment deletes from crashing deployment lookups
delete_deployment and upsert_deployment shrink model_list before they repair
the index maps, so a lookup on another thread in that window read a stale
index: IndexError, or the neighbouring deployment and its credentials. An add
racing a delete could corrupt the index for good.
One RLock now covers model_list and the three index maps. Delete, upsert, add
and the set_model_list reset change them under it, and every lookup that turns
an index into a deployment reads under it.
Fixes#45242
* test(router): release the held delete when the read reaches the index lock
The race tests released the held write after a fixed 0.2 s, so a reader the runner
stalled past it read the repaired list and passed on unfixed code. The read now runs
on its own thread and the write stays held until the read returns or reaches the
router's index lock; the 10 s bound only fails a broken synchronization.
The integration test gains the audit cells: chat, streaming, Responses and
Anthropic messages calls (sync and async, by model name and team public name)
during a held delete, the same for upsert, add and set_model_list, a request for
the deployment being deleted, a provider error, a double delete, and thread and
async bursts during upsert churn.
* test(router): cover the early return of a delete and the team filter of get_model_ids
delete_deployment now returns None from inside the index lock, and get_model_ids by
name reads its deployments through _deployments_named; pin both: deleting an unknown
id returns None and leaves the list alone, and exclude_team_models drops only the
team deployment.
* fix(router): read deployments by name under the index lock outside the Router
The response polling provider check, the complexity router's adaptive pool and the
savings baseline turned the name index into list positions themselves, so a concurrent
delete could still hand them a stale slot. They now read through Router.deployments_named,
which takes the deployment index lock.
* fix(router): type the locked deployment lookups as DeploymentTypedDict
deployments_named and _team_deployments_named returned tuple[Any, ...]. They now return
DeploymentTypedDict rows, which is what model_list holds. The adaptive pool setup in the
Router and the complexity router drop their branches for pydantic Deployment rows, which
these lookups never return, and should_include_deployment accepts the typed row.
---------
Co-authored-by: yucheng-berri <293926326+yucheng-berri@users.noreply.github.com>
* fix(ui): keep Aria accessible in the dashboard header
* test(ui): wait for every pagination fixture log
* test(ui): update assistant setup for Aria branding
* fix(guardrails): exclude litellm_metadata from pass-through guardrail scan text
The pass-through guardrail translation json-dumps the request body as the
scan text when no request_fields are configured. The unified guardrail hook
injects litellm_metadata into that body first, so the dump carried the
caller's key metadata. Exclude it alongside metadata and litellm_logging_obj;
vendors still receive it through request_data.
* test(guardrails): type the pass-through scan-text tests per AGENTS.md
* fix(otel): decide per-team Arize sampling from the trace id so every pod agrees
* test(otel): cover per-team sampling across pods, workers, endpoints and restarts
* fix(otel): remember only a failed tree's sampling verdict so a later failed call answers to the error rate
A successful tree's verdict follows from the trace id alone, so remembering it only made a later failed call of the same trace follow the success rate on the pod that served both, while other pods judged it at the error rate
* test(otel): flush every worker before the restart cell kills it and cover a failed call after a dropped one
* fix(otel): decide team sampling per request, not per trace
Calls sharing a caller's trace id are separate requests, keyed by their root span.
Each is decided by the trace hash at the rate for its own outcome, and only its own
late spans follow its verdict, so no pod applies an earlier call's verdict to a later
call of the trace
* fix(otel): forget a span's request root when it ends, not when newer spans start
The root of each open span was kept in a first-in-first-out table of 16384 entries,
so a span two levels below the root that stayed open while the worker started more
spans than that lost its entry and ended under its parent's id. Its request then
never closed and was held from every sampled destination until shutdown.
Each open span now keeps its entry until it ends, and a span starting under a parent
that has already ended reads the root from the parent span itself
* fix(otel): decide a request as its spans end once it falls out of the open span table
A request whose span was pushed out of the table of open spans, because the worker
held more than 16384 open at once, ended that span under another key and was held
from every sampled destination until shutdown. Pushing a span out now drops its
request's open count too, so the request is decided as each of its spans ends, the
way one pushed out of the open request table already is
* test(otel): annotate the team sampling test locals as Final
* test(otel): build the caller's remote parent in one helper instead of rebinding it per trace
* fix(otel): keep each open span's request until it ends, bounded by the open requests
The span table had its own bound of 16384, so a worker holding more spans open than
that pushed out the oldest, and a failure in such a span no longer reached its
request. Each request now holds the ids of its open spans, and a span's entry in the
lookup goes when the span ends or when its request is pushed out of the open request
table, the same bound main already applies. A failure anywhere in a tracked request
reaches its verdict however many spans are open
---------
Co-authored-by: yucheng-berri <293926326+yucheng-berri@users.noreply.github.com>
* fix(streaming): stop re-raising and re-logging a failed stream on every next()
A stream that raised keeps its provider iterator alive, so a consumer that iterates past the error re-detected, re-raised and re-logged the same failure on every call, flooding failure callbacks for one request. Latch the failure and re-raise it without logging again.
* fix(streaming): drop the retained traceback before re-raising a failed stream
* test(streaming): count failures through a real logging callback and require a clean stop on later reads
* test(streaming): require the stuck stream to close early and release its socket
The provider stream now records a GeneratorExit and caps its chunks, so draining it or detecting the loop late fails. Two wire cases check a sync and an async stuck stream release their TCP session.
* test(streaming): avoid failure recorder name collision
---------
Co-authored-by: Devin AI <devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
Co-authored-by: moe-berri <moe@berri.ai>
Preserve oversized JSON integers in proxy requests and model rewrites
Parse large bodies in a worker thread to avoid blocking request handling
Fixes#47947
Co-authored-by: Jeevan Babu Gotru <114808955+jeevan1098@users.noreply.github.com>
Co-authored-by: moe-berri <moe@berri.ai>
* fix(deepseek): pass reasoning_effort through to the API
DeepSeekChatConfig.map_openai_params mapped reasoning_effort to a binary
thinking {type: enabled/disabled} toggle and discarded the effort value, so
DeepSeek always ran default high effort. High-effort reasoning can run away
and exhaust max_tokens during reasoning, returning empty content. Preserve
the reasoning_effort value into the request body when thinking is enabled so
callers can bound effort (e.g. reasoning_effort="low"). Backward-compatible.
See https://api-docs.deepseek.com/guides/thinking_mode
* test(deepseek): reasoning_effort passthrough coverage
* fix(deepseek): exclude reasoning_effort=none sentinel from passthrough
"none" is the thinking-OFF sentinel, not a real effort value. Passing it
alongside explicit thinking:enabled would send contradictory instructions to
DeepSeek. Only forward real effort values (low/medium/high/xhigh/max).
Addresses review feedback on the reasoning_effort passthrough change.
* refactor(deepseek): name supported effort values and type mapped params
* fix(deepseek): keep mapping types local to the adapter
* fix(deepseek): remove cast and type mapped return
---------
Co-authored-by: moe-berri <moe@berri.ai>
* fix(proxy): reconcile team member total_spend with current cycle spend
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): read /team/list as a list and drain the scripted provider
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(repositories): compose the membership reset writer and drop SQL text asserts
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: type the cascade repair test and flatten memberships without a stacked comprehension
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: keep the reset job test file's existing typing untouched
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(cost): parse 12-hour cache write rates for batch and ultrafast tiers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(cost): bill 12-hour batch cache writes once when only the 12hr batch key exists
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(teams): show caller member spend and budget on Your Teams
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(teams): use member-linked reset dates
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(teams): keep type-discipline suppressions on cast lines
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(teams): preserve empty page list response
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(teams): keep the member reset date inside the spend column
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(teams): match email-only roster members for caller membership
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show team member budget reset date in UTC
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(teams): address review on caller membership enrichment
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(ui): label the caller's team spend line My Spend
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(opencode): bill OpenCode Go requests from the reported list_cost
OpenCode returns usage.opencode.list_cost on every response. Read it into the provider-reported cost the cost calculator checks before the price map, for chat and Responses, streamed and not. Keep the static rows as the fallback, and stop the generic usage.cost path from billing Go requests at 0 (usage.cost is what hit the prepaid balance, 0 inside the plan).
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(core): resolve reported cost breakdowns without recursion
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(opencode): cover reported list_cost spend through the proxy
Replaces the recursive _reported_cost_value with a flat _cost_number helper
so the code-quality gate passes, and adds an integration test that runs the
proxy against a scripted OpenCode upstream for a non-stream and a stream
chat completion, asserting the response cost header and the spend log row
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(opencode): bill native /v1/messages from the reported list_cost
The native Anthropic Messages route priced OpenCode Go bodies from the price
rows twice over: the non-stream body is a TypedDict the response-cost header
path priced as-is, and the stream assembled its ModelResponse with the rows
cost already stamped. Both now read usage.opencode.list_cost, the non-stream
body through billable_result and the stream from the final message_delta.
The integration test now covers chat, responses and messages, stream and
non-stream, through the proxy against a scripted OpenCode upstream
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(opencode): keep the stream cost stamp inside the lint gate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(opencode): read list_cost off dict chunks and trust it only from OpenCode on /v1/messages
Stream builders receive chunks both as mappings and as ModelResponseStream
objects, so the list_cost scan now reads usage through chunk_usage. The
native Messages route only trusts a reported cost from opencode_go, every
other upstream keeps the price map, which also keeps a zero usage.cost from
being billed as free. Served bodies without stop_sequence still price, and
the new code stays inside the basedpyright gate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(proxy): add disable_pre_routing_error_logs to skip spend rows for pre-routing rejections
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): decide pre-routing rejections by pipeline stage, not exception type
Guardrails such as AIM raise ProxyException(400), the same shape as an unparsable
body, so the status code list skipped their spend rows. The pipeline stamps
litellm_call_id only after auth, body parsing and model resolution, so a failure
reaching the hook without one died before any provider was considered.
_handle_logging_proxy_only_error flags those requests when it mints the call id,
and is_pre_routing_rejection keys off that flag plus ProxyModelNotFoundError.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): mark pre-routing rejections with a server-owned sentinel
request_data is the client body, so a boolean marker or a litellm_call_id in it
can be sent by the client to hide a guardrail block or force a spend row. The
failure handler now stores an object() sentinel whenever it finds no logging
object, and is_pre_routing_rejection checks identity, which JSON cannot forge.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): skip declared text-only deployments for image requests without enable_pre_call_checks
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): run the vision filter after the cooldown filter so a cooled vision leg falls back to the text leg
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): apply the vision filter after pre-call checks and tidy the new tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(router): mark the remaining locals in the vision filter tests Final
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(guardrails): add native Reco guardrail provider (pre_call)
Registers guardrail: reco as a first-class vendor integration, built on
the existing generic guardrail API wire contract rather than reimplementing
the HTTP client and payload mapping. Maps reco_tenant_id to the
X-Reco-Tenant-Id header and hardcodes fail_open/fail_on_error since those
should not be customer-configurable. v1 only supports mode: pre_call;
post_call and streaming redaction already work against the existing
generic pilot and are deferred.
Also overrides apply_guardrail to re-raise blocked-request errors with the
guardrail's own configured name: the base class hardcodes a module-level
name in that exception, which would otherwise mislabel every blocked Reco
request in logs.
* feat(guardrails): configurable Reco fail modes
Expose unreachable_fallback and fail_on_error on the Reco guardrail
and keep fail-open as the default. Values under optional_params win.
Top-level litellm_params values apply only when set explicitly, since
their defaults are fail-closed and would otherwise flip the default
Also type RecoGuardrail's forwarded kwargs with an Unpack'd TypedDict
instead of silencing ANN003
Signed-off-by: MickeyShnaiderman-RecoLabs <mickeys@reco.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
* chore(guardrails): align Reco with main's CI gates
Upstream moved unit tests from tests/test_litellm to tests/unit, so
test_reco.py was invoked by no CI job. Move it next to the other
guardrail tests
Replace RecoGuardrail's forwarded kwargs with explicit keyword
params, read optional_params through a pydantic adapter instead of
untyped dict/getattr access, freeze RecoOptionalParams, and drop
mutable-ok markers that suppress nothing, to stay within main's
type-discipline and basedpyright budgets
Signed-off-by: MickeyShnaiderman-RecoLabs <mickeys@reco.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
* feat(guardrails): fail closed by default in the Reco guardrail
Matches GenericGuardrailAPI and every other guardrail here: a Reco error or outage now blocks the request unless the admin sets fail_on_error: false. Tests cover the default, explicit fail-open, and top-level vs optional_params precedence
Co-authored-by: RecoLabs-YaronBratspiess <208601538+RecoLabs-YaronBratspiess@users.noreply.github.com>
Co-authored-by: MickeyShnaiderman-RecoLabs <77006830+MickeyShnaiderman-RecoLabs@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(guardrails): honor explicit fail-open on Reco when optional_params is a model
model_dump(exclude_unset=True) so pydantic defaults in optional_params no longer override an explicit top-level fail_on_error. Also drop the duplicate log_guardrail_information decorator, call the parent by keyword, freeze RecoConfigModel, and type the tests
Co-authored-by: RecoLabs-YaronBratspiess <208601538+RecoLabs-YaronBratspiess@users.noreply.github.com>
Co-authored-by: MickeyShnaiderman-RecoLabs <77006830+MickeyShnaiderman-RecoLabs@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(guardrails): keep Reco config model assignable and the hook decorator the gate requires
frozen=True on RecoConfigModel froze the shared LitellmParams mixin, which the guardrail registry assigns resolved secrets on, so every guardrail failed to load. Dropped it and added a test that LitellmParams stays assignable. Restored log_guardrail_information on the Reco override since tests/code_coverage_tests/check_guardrail_apply_decorator.py requires it on every hook. Test params typed as Mapping for LIT001
Co-authored-by: RecoLabs-YaronBratspiess <208601538+RecoLabs-YaronBratspiess@users.noreply.github.com>
Co-authored-by: MickeyShnaiderman-RecoLabs <77006830+MickeyShnaiderman-RecoLabs@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(guardrails): format the Reco test file
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Signed-off-by: MickeyShnaiderman-RecoLabs <mickeys@reco.ai>
Co-authored-by: Yaron Bratspiess <yaronb@reco.ai>
Co-authored-by: MickeyShnaiderman-RecoLabs <mickeys@reco.ai>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: MickeyShnaiderman-RecoLabs <77006830+MickeyShnaiderman-RecoLabs@users.noreply.github.com>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(bedrock): price claude_platform calls from the anthropic cost map rows
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix: avoid recursive Claude Platform lookups
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(bedrock): price claude_platform calls from their own cost map rows
Replace the claude_platform -> anthropic lookup redirect with
bedrock/claude_platform/* rows, and apply the inference_geo multiplier
in the bedrock cost path so US-only inference is priced at 1.1x
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* test(bedrock): exempt claude_platform rows from the Converse routing guard
The claude_platform/ route is explicit in the row key, so these rows are
not accidental non-Converse Bedrock chat rows
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(anthropic): stop injecting empty-content placeholder into tool-call-only assistant turns
_sanitize_empty_text_content replaced empty or whitespace-only text with
"[System: Empty message content sanitised to satisfy protocol]" for every user and
assistant turn. anthropic_messages_pt calls it unconditionally, independent of
modify_params, because an empty text block always makes Anthropic 400.
An assistant turn carrying tool_calls and no text is the normal shape an OpenAI-format
agent emits, and Anthropic accepts an assistant turn whose content is just tool_use, so
there was nothing to substitute there. Substituting anyway put a protocol diagnostic in
the content field, which is the assistant's utterance as far as the model and the client
are concerned, so it got persisted into conversation history and echoed back on later
turns.
Empty text is now dropped rather than replaced when the assistant turn already carries
substantive content (tool_calls, or non-text blocks such as tool_use). Turns that would
otherwise be left with no content at all keep the placeholder, since Anthropic rejects
those too.
Fixes#24498
* fix(anthropic): only drop empty text when a sibling block survives conversion
_carries_non_text_content treated any non-text block as substantive, but the assistant
list-content converter only keeps thinking (when signable), server_tool_use and
*_tool_result blocks. For an assistant turn holding empty text next to a block it
discards, such as an image_url or a signature-less thinking block, dropping the empty
text left nothing behind and the whole turn disappeared from the request.
The predicate now mirrors what the conversion actually keeps, so those turns retain the
placeholder and survive. This also improves on the previous behaviour, where such a turn
was dropped outright.
* test(anthropic): cover malformed content blocks in the survivor predicate
_survives_assistant_conversion had a separate non-dict guard that no test reached,
since _is_empty_text_block already rejects non-dict blocks before the predicate runs
on them in practice. Folding the check into the type lookup keeps the same behaviour
for bare strings and blocks without a type key, with no unreachable branch.
Extends the sibling-block test with a block missing its type key.
* test(anthropic): set modify_params through monkeypatch in sanitization tests
The test-quality gate flags direct writes to litellm module globals. Swap the
class setup/teardown for an autouse monkeypatch fixture, drop writes that just
repeat its default, and route the False cases through monkeypatch
---------
Co-authored-by: tbartik <22518863+tbartik@users.noreply.github.com>
Co-authored-by: ryan-crabbe-berri <ryan@berri.ai>
* fix(vector_stores): put azure vector store id before api-version query
Azure OpenAI's get_complete_url returns a url ending in ?api-version=...,
and the vector store handlers appended the id with an f-string, producing
.../openai/vector_stores?api-version=X/vs_abc and a 404 from Azure.
Add BaseVectorStoreConfig.get_vector_store_url, defaulting to the previous
behaviour so other providers are unchanged, and override it in
AzureOpenAIVectorStoreConfig to keep the query string last. The six
retrieve/update/delete handlers now call it.
Co-Authored-By: Claude <noreply@anthropic.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(vector_stores): trim comments on get_vector_store_url
Drops the inline comment in the Azure override and shortens the base docstring, per AGENTS.md
Co-authored-by: Keith-Hardaway <276775292+hardawaykwork@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(vector_stores): capture the Azure retrieve request without a mutable url list
Greptile on #48224: the handler test now follows the file's received list pattern with Final bindings
Co-authored-by: Keith-Hardaway <276775292+hardawaykwork@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Keith-Hardaway <keith.hardaway@docusign.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Keith-Hardaway <276775292+hardawaykwork@users.noreply.github.com>
* fix(router): keep latency routing alive when latency samples are unusable
latency-based-routing ranks deployments by the average of the cached samples
and keeps everything within lowest_latency_buffer times the lowest. A negative
or NaN sample breaks that invariant, because the minimum no longer satisfies
its own bound. The candidate list then comes back empty and random.choice
raises IndexError, which turns every request to the group into a 500 until the
cache entry expires.
Samples come from completion_start_time - start_time, and that difference is
negative whenever the two timestamps invert. Production recorded samples
around -600s from successful calls.
Drop non-finite and negative samples before ranking, and fall back to every
candidate if the window is empty anyway, for example with a negative buffer.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Co-Authored-By: AaronHowell <111272993+AaronHowell@users.noreply.github.com>
* test(router): seed latency_normalized in the unusable TTFT fallback test
Main reads latency_normalized since #47727, so seeding latency left both deployments ranked at zero and the test picked at random.
Co-authored-by: Rick <26716961+Bytechoreographer@users.noreply.github.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(router): assert ranked latency after dropping unusable TTFT samples
Greptile on #48216: the new tests passed with the sample filter removed because the empty-window fallback alone saved them. They now assert _latency_per_deployment so unfiltered negative or NaN samples fail deterministically. Also adds Final to ttft_samples and drops a comment that restated the code
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(router): drop Final on a loop-local in latency ranking
basedpyright rejects Final on a variable assigned inside a loop, and the type_check_gate counts it as a new reportGeneralTypeIssues error
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(router): type the request kwargs in the unusable TTFT tests
Greptile on #48216: bare dict annotations. The tests now hold the metadata dict in a typed local and assert on it directly
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Rick <26716961+Bytechoreographer@users.noreply.github.com>
Co-authored-by: Claude Code <noreply@anthropic.com>
Co-authored-by: AaronHowell <111272993+AaronHowell@users.noreply.github.com>
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(typing): replace pass-through Any with object in 13 annotations
* docs(client): say dict[str, object] in the models get docstring
* refactor(typing): drop the unused Span aliases in base_cache and levo
* refactor(typing): take a Mapping in delete_nested_value and cover the fresh copy
* fix(typing): deep-copy a read-only Mapping in delete_nested_value through a dict
