* ci: warn on SQL IN lists with no written bound
Postgres caps a prepared statement at 32,767 bind parameters and a
membership filter binds one per value, so an IN list built from table
data breaks once the table outgrows the cap. That is how the budget reset
job froze every due budget (LIT-7535, #40564).
check_unbounded_in_lists.py reports every Prisma "in" / "not_in" filter
whose value has no fixed size and every raw SQL literal that splices a
list in after "IN (", unless the line carries "# bounded-ok: <reason>".
It only warns for now: the output is the inventory for RCA action item
AI-1, and it exits 0.
* ci: decide a constant IN list by its module binding, not its casing
An ALL_CAPS name imported or filled at runtime is as unbounded as any
other, so a name now passes only when the module binds it once to a
value of fixed size. Adds Final to the locals a loop does not forbid.
* ci: only a frozen module value makes an IN list constant
A module list bound once could still grow through append or extend, so
a name now counts as fixed only when it is bound to a tuple, frozenset
or constant. Trims the module docstring to what a reader needs.
* ci: chunk Prisma IN lists with a shared helper and fail on new unbounded ones
Add litellm.repositories.bounded_in: find_many_in, count_in, update_many_in
and delete_many_in split a deduplicated value list into 5,000-value chunks,
AND each chunk with the caller's where, run them in order (a transaction
handle works) and combine the results. Writes take a required atomicity
argument, and a where that already filters the chunked field is refused.
check_unbounded_in_lists.py now fails CI on any finding missing from
unbounded_in_baseline.txt and on any stale baseline entry, so the baseline
only shrinks. Entries are keyed by path, enclosing scope, kind, field and
occurrence, not line numbers. The helper module is exempt, a constant
spread into a frozen tuple counts as fixed, and messages point at the
helper for "in" and at an array parameter for "not_in" and raw SQL.
A real-Postgres integration test shows a raw 40,000-value filter rejected
for too many bind variables while the helpers handle it.
* refactor: rename bounded_in to chunked_in and let callers pick a chunk size
The helper module is litellm.repositories.chunked_in, and its unit and
integration tests, the checker's exemption path and its finding messages
follow the new name. The `# bounded-ok` marker is unchanged.
find_many_in, count_in, update_many_in and delete_many_in take a
keyword-only chunk_size, defaulting to IN_LIST_CHUNK_SIZE (5,000). A value
below 1 or above MAX_IN_LIST_CHUNK_SIZE (30,000) raises ValueError before
any query, which leaves the rest of the filter headroom under Postgres's
32,767 bind-parameter cap.
* refactor: flatten chunked_in's stacked comprehensions with chain.from_iterable
LIT014 (#42650) caps a comprehension at one for and one if clause. The four nested walks in the helper now chain their iterables instead, with the same order and results.
* refactor: recover user details with find_many_in, sending chunks as lists
_details_for_user_ids reads users through find_many_in instead of a raw
"in" filter, so its lookup stays under the bind-parameter cap for any
number of recovered keys. Up to 5,000 ids it still sends one find_many
with the same where dict, and a PrismaError from any chunk is still
logged and treated as no details.
The helper now sends each chunk as a list, so a chunked filter equals
the dict a hand-written call would send and a migrated call site's
existing assertions keep passing.
The site's baseline entry is gone.
* ci: skip functional TypedDict field maps in the unbounded IN list check
The dict passed as the field map of TypedDict("Name", {...}), or as its fields= keyword, names fields: an "in" or "notIn" key there is a type, not a filter. Only that dict is skipped, for TypedDict, typing.TypedDict and typing_extensions.TypedDict; a filter nested in a field value or passed to any other call is still reported. The two types/proxy/management_endpoints/team_endpoints.py entries leave the baseline, which is now 156.
* fix: refuse an update_many_in whose data writes the chunked field
Chunks run one after another, so an update that sets the chunked field can move a row into a later chunk, which updates it again and counts it twice: values ["old", "new"] with chunk_size=1 and data={"id": "new"} does exactly that. update_many_in now raises ChunkedFieldWriteError before any query when data has the chunked field as a top-level key, in any form, including Prisma operators such as {"set": ...}.
* docs: cut the unbounded IN list checker's docstring to what it flags and how to clear it
It now says what is reported, the three ways to clear a finding, and how the baseline and --update-baseline work, in 11 lines. The per-shape detail lives in the tests.
* ci: key an unbounded IN list finding by its filtered expression too
A baseline key of path, scope, kind, field and occurrence let a PR delete
a baselined filter and add a different unbounded one on the same field in
the same function, and the new one took over the old key. The key now
also carries the filtered expression's source, whitespace-normalized
(the Prisma value, or a raw-SQL `IN (...)` slot), so that swap reads as
one new and one stale entry and fails the run. The same expression
re-added in the same function is still the same finding.
Every baseline entry is rewritten in the new form; the 156 findings are
unchanged, and only occurrence indexes renumber where one field had
several different expressions.
Register Sail (providers.json, LlmProviders.SAIL, OpenAI-compatible lists,
ProviderConfigManager) for chat, Responses and /v1/messages, and add its 12
models to both cost maps with asap, balanced and flex price columns.
Sail picks speed and price with metadata.completion_window and rejects
service_tier, so the Sail chat and Responses configs translate the tier:
default and priority to asap, flex to flex, balanced to balanced, auto to no
window. Billing prices the window that was sent. A tier Sail has no window
for, or a window or tier set where billing cannot see it (request metadata,
extra_body), is a 400 unless drop_params is set.
Add balanced to ServiceTier and its _balanced price columns to the model
info types, the Rust catalog and the dashboard schema. A transform_extra_body
hook on the chat and Responses base configs, which returns extra_body
unchanged by default, lets Sail keep the window when a caller also sends
extra_body.metadata. Sail is listed in the Add Model form and model picker.
Co-authored-by: shrey kharbanda <shrey@berri.ai>
* fix(guardrails): block private destinations in custom code http_request and bound guardrail execution time
* fix(guardrails): keep startup fail-closed on a custom code compile error and report a load timeout on the test endpoint
A compile failure is no longer a ValueError, so a config-file custom code guardrail that does not compile
stops the proxy at startup as it did before, while POST /guardrails catches it by name and still rolls back.
The admin test endpoint reports a module-level timeout as an execution timeout instead of a compile error,
a caller-supplied Host header is stripped from http_* requests while validation is on, and GET keeps the
shared client's connect timeout.
* test(guardrails): cover the http_request methods, header passthrough and cancellation paths
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(responses): run stream failure and success hooks on the iterating loop instead of blocking it
A dropped provider stream on native /v1/responses ran the failure logging through run_async_function from inside the async iterator, which parks the event loop thread on a helper-thread future until every failure callback returns, and never returns when a callback waits on state only that loop can advance. With a running loop the failure handlers (and the completed-stream success deployment hook) are now scheduled as tasks on it, the way chat streaming already does; the sync iterator keeps its blocking path
* fix(responses): await stream failure and success logging on the iterating loop before propagating
Keep the merge-base hook set for the native Responses stream: async_failure_handler plus the executor-thread failure_handler on failure, and the post-call success deployment hook on completion. Inside a running loop the async handler is scheduled as a task on that loop and the async iterator awaits it before re-raising, so the loop is never blocked on a foreign-loop future and the failure is attributed before the router's fallback wrapper re-enters the same logging object. The sync iterator inside a running loop keeps the task fire-and-forget with a strong reference.
* fix(responses): submit the sync failure handler only after the async one finishes
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* feat(proxy): add fail_closed_rate_limit_enforcement to reject requests with 503 while Redis rate limit counters are unreachable
* fix(proxy): reject fail-closed rate limit checks before logging the in-memory fallback and pin the boot warning in the lifespan
* fix(proxy): coerce the fail-closed flag, fail closed on read-only checks, and refund partial cluster increments
* fix(proxy): window-guard rate limit refunds and catch the fail-closed rejection by type
* fix(proxy): read the compaction rate-limit gate's limiter from the proxy hook registry
* fix(proxy): count the pending request in read-only rate-limit checks and keep the compaction gate off the caller's parallel slot
The compaction polyfill's summary-model gate, once it ran against the real v3 limiter, showed two behaviors nobody had chosen. The read-only check compared the stored counter with the same `>` the increment path uses, but a read-only check decides a request that has not been counted yet, so a summary model exactly at its rpm limit still went out. The read-only path now adds the pending increment of 1 before comparing; the increment path is unchanged.
The gate also passed the key's max_parallel_requests gauge through, and the read-only gauge count includes the caller's own in-flight slot, so a key with max_parallel_requests: 1 never compacted. The gate now drops that gauge from its descriptors, since the summary call runs inside a request the limiter already admitted.
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* test(integration): pin the team-admin status-code matrix across every management door
Every endpoint that admits a team admin today is called as a proxy admin, an admin of the
target team, a plain member, an admin of another team and a teamless user, and the current
status code is asserted per actor. The matrix is the parity check for collapsing the five
team-admin helpers into one shared gate and for the later default-off permission flip.
* test(integration): pin the permission-enabled team-admin doors in the gate matrix
Adds three doors that run with team_admin_editable_team_fields granting max_budget, projects and member_key_budgets, so the enabled path is pinned alongside the default-off one. Hoists the ui_settings toggle from test_warmed_policy into the shared client so both files use one helper
* test(integration): grant each permitted door only the permission it needs
A door now names its single grant instead of every permission at once, so a gate that checks the wrong permission for a route turns that door red
* test(integration): rewrite the team-admin matrix rows as request plus expected codes
Each row now names the route it calls and the code each caller gets, and creates the member, key, model, callback or invitation it acts on through plain helpers on the shared team. Drops the Need, Target, World and Door types and the prepare step that seeded fixtures by enum.
* fix(otel): detach post-response service spans by request phase, name redis spans by operation
Service spans logged from the post-response phase (success callbacks, the response-cache write) now root their own trace linked to the request span even while the server span is still recording, instead of only when they happen to end after it. Redis service spans are named `redis <operation>`; the litellm call chain that issued them moves to the `litellm.service.caller` attribute via a typed `ServiceLoggerPayload.caller` field.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep the service caller on failure and legacy spans, test the production phase dispatch sites
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): mark anthropic messages stream cache write as post-response phase
The /v1/messages streaming cache writer awaits async_add_cache inline
instead of going through create_cache_write_task, so its redis span
stayed parented under the request trace. Wrap the write in
post_response_phase so it detaches like the chat completions write.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(anthropic): write the Messages stream cache in a background task after handoff
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yassin <yassin@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ci): stop five stale or flaky CI reds and retry CyberArk policy-load conflicts
The Langfuse redaction unit test exports to a local OTLP capture instead of
polling Langfuse Cloud through a recorded lookup. The passthrough worker-kill
test only requires spend rows for requests the surviving worker served. The
spend-routes sweep treats the intentional /spend/capture_rate 503 as expected.
CyberArk retries a 409 policy load in Python, Rust and the e2e Conjur helper
instead of reading it as "variable exists". The integration egress guard now
matches the script's own cgroup, so it no longer blocks the CircleCI agent,
which runs as the same user.
* fix(ci): keep the policy-load backoff typed as float
* fix(ci): retry CyberArk policy loads without blocking the event loop and tighten the worker-kill and Langfuse tests
* fix(secrets): load CyberArk policy one request at a time per manager
* test(secrets): pin that non-conflict CyberArk policy failures are not retried
* test(unit): run tests/unit with only an allowlisted host environment
CircleCI's unit job inherits every project env var, so real provider keys,
REDIS_HOST, DATABASE_URL and AWS or Azure credentials reached tests that
assume none are set. Locally, litellm's import-time load_dotenv did the same
from any .env up the tree. The unit conftest now drops every variable outside
a small allowlist and disables dotenv before litellm is imported.
* test(e2e): name a failed search and the stuck batch status instead of misattributing them
The websearch session test read an empty web_search_tool_result_error block as a
successful search, so a failing search tool surfaced as a session billing bug.
The batch cancellation timeout now reports the last status the proxy returned.
* fix(ci): scrub the host environment per unit test instead of for the whole pytest process
GHA shards run tests/unit next to other suites in one process, so the import-time
scrub deleted MCP_TEST_PEER_PYTHON before tests/mcp_tests read it and the MCP
upstream fell back to the SDK2 interpreter. The two websearch tests that called
OpenAI and Perplexity live are removed: tests/unit no longer sees their keys.
* fix(ci): scrub only the host variables present before litellm is imported
The per-test scrub also deleted TIKTOKEN_CACHE_DIR, which litellm sets at import to
its bundled encodings, so tokenizer paths tried to download them and hit the
socket guard. The prisma setup test now passes its own database URL instead of
reading one another test leaked into the process environment.
* fix(ci): stop the order-dependent unit reds and settle logging tasks on their own queue
LoggingWorker marked a task done on whichever queue was current when the callback
finished, so a callback that outlived an event-loop change raised "task_done()
called too many times" or undercounted the new loop's queue. It now settles the
queue the task came from.
The rest are test isolation fixes for failures that only appeared when another
file ran first on the same xdist worker: a replaced user_api_key_cache, breaker
metrics unregistered by prometheus tests, semantic_router's health-check filter on
uvicorn.access, logging tasks carried over from bedrock tests, a Router-written
model_cost entry, and a stray post captured by the langflow test. The token
counter check now asserts bounded chunking instead of wall-clock time.
* test(e2e/ui): wait for the logout redirect before visiting a protected page
Logout revokes the session server-side before clearing cookies and navigating, so an immediate page.goto either ran with the cookie still set or was aborted by the logout redirect (net::ERR_ABORTED).
* test(unit): restore the prometheus metrics config per test and settle logs carried from earlier tests in the a2a cost tests
* test(router): pin the router clock in the usage counter tests so a minute rollover cannot empty the read
* test(e2e/ui): wait for logout to clear the token cookie instead of for a login redirect
* test(integration/mcp): answer the model-info probe another test's proxy sends to the model double
