lint: the real cause was never F401 (that step already passed) - it
was scripts/test_quality_gate.py's TQ003 rule, which flags
sys.path.insert as unnecessary since pytest's rootdir handling and the
installed package already make litellm importable. Removed it along
with the now-unneeded os/sys imports and the __main__ block that
depended on it, matching test_scx_ai_provider.py's cleaner convention.
Confirmed directly: scripts/check_test_quality.py now reports 0
violations on this file (it still reports 2 on test_xiaomi_mimo.py,
the template this was built from - grandfathered only because it
predates the gate).
code-quality: check_provider_folders_documented.py requires every
litellm/llms/openai_like/providers.json entry to have a matching entry
in provider_endpoints_support.json. sagg had none. Added one,
chat_completions only (the only endpoint actually verified in this
PR's own proof of fix) - confirmed the check now passes, and fails
again with the exact expected error when the entry is removed.
osv-scan: not caused by this PR. Zero dependency/lockfile files are
touched anywhere in this diff. Cross-checked two other open,
unrelated PRs against the same base branch: one also failed osv-scan
in the same time window this PR's check ran, the other two (scanned
~4.5 hours later) both passed - a transient, time-bound,
repo-wide/base-branch issue, not something this PR introduced or can
fix.
Remove unused MagicMock/patch imports (F401, blocking required lint
checks). Remove the live-network completion test: tests/e2e/CLAUDE.md's
Hard Rules forbid substituting a unit test for e2e feature coverage,
and this env-var-gated call doesn't fit the e2e harness either (no
proxy, no ProxyClient, no coverage-registry marker) - a proper e2e
addition is a separate, larger piece of work, out of scope here. The
real live proof stays in the PR description's Proof of Fix section,
where CLAUDE.md says it belongs. Trim docstrings and inline comments
that only restated the assertion on the next line, per CLAUDE.md's
comment policy - kept the two that explain non-obvious behavior
(first-slash-only provider splitting, why the param mapping matters).
SAGG (https://api.privatedeskai.com) is an OpenAI-compatible LLM
inference gateway with automatic multi-provider failover. Adds it as a
dynamic JSON-registry provider: base_url + api_key_env in
providers.json, plus the required LlmProviders enum entry so it
resolves via get_llm_provider() and appears in litellm.provider_list.
param_mappings.max_completion_tokens -> max_tokens is required, not
optional: SAGG's own request parser only recognizes max_tokens (has no
max_completion_tokens field at all), so a caller using the newer OpenAI
param name would otherwise have it silently dropped.
Verified end-to-end against the live SAGG API, including that the
max_completion_tokens mapping is actually enforced server-side
(finish_reason: length at the requested cap), not just unit-tested.
* fix(auth): quiet malformed virtual key rejections to stdout
Reduce noisy invalid-api-key error logs by classifying malformed virtual
keys and routing their rejections to stdout as WARNING instead of stderr
as ERROR. Suppressible via LITELLM_LOG=ERROR or log_client_error_tracebacks=true.
Changes:
- auth_utils: is_invalid_virtual_key_error() classifier and marker functions
- auth_exception_handler: log invalid keys as WARNING to child logger before
identity seeding and callbacks, escalate non-401 transforms to ERROR
- user_api_key_auth: websocket early-raise WebSocketException(1008) to avoid
double-logging at HTTP layer
- _logging: child logger verbose_proxy_stdout_logger with no handler/level;
LevelRoutingStreamHandler routes its WARNING records to stdout; handler
setLevel in _turn_on_json() closes JSON config handler level leak
- test_auth_exception_handler: new test case verifying malformed-key logs
at WARNING with marker retention through transformations
Fixes LIT-5362
* fix(auth): classify malformed-key 401 by raise-site marker, not message text
Review round 1 (Greptile P2, veria Low):
- Move the marker attribute name to litellm/constants.py per the shared
sentinel convention
- Stamp the marker on the malformed-key 401 where it is raised and classify
only by it. Message text is caller-influenceable on other 401s (vector
store ids, organization ids are interpolated into their messages), so a
phrase match would let a request body demote an authorization failure to
the quiet log path
- Regression test: a 401 carrying the phrase but not the marker stays at
ERROR on stderr
* fix(ui): keep litellm_credential_name from LiteLLM Params JSON when no credential is selected
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(ui): drop null litellm_credential_name from AddModelPanel payload fixture
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>
ProxyExtrasDBManager.spend_logs_is_partitioned() (#38452) silently returns
False when psycopg can't be imported, and psycopg was never added to the
extra_proxy install, so every production image lacks it. Schema
reconciliation then generates the unfiltered primary-key rewrite against a
genuinely partitioned LiteLLM_SpendLogs and Postgres rejects it, exactly the
failure the fix was meant to prevent. Ships psycopg via extra_proxy and logs
a warning when it's still missing instead of failing silently.
At end of drain the pump enqueued the sentinel first and picked the
billing mode from client_detached afterward, so a client that consumed
the sentinel and tore the relay down before the pump resumed (possible
whenever the sentinel enqueue hit a full queue) had its fully delivered
response billed through the teardown path, skipping the proxy's
post-response hook. Bill or park before the sentinel goes out, and let
an unconsumed sentinel fall back to dispatching the parked billing.
Shadow eval jobs previously targeted only virtual keys, so deployments on
pure JWT auth (which present no key at all) could never sample their
traffic. Jobs now carry a typed (target_type, target_id) pair covering
keys, teams, and users; sampling matches the identity every request
resolves to at auth time, so team and user jobs cover JWT traffic with
no client changes.
Resolves LIT-6578
* feat(complexity_router): escalate oversized prompts to a tier that fits before dispatch
The classifier scores complexity and never prompt size, so a long agentic
session whose newest ask is trivial classifies SIMPLE onto a small-window
tier and the provider rejects it with a context-window 400 that nothing
retries. The gate runs after classification on every decision path
(classify tail and session-affinity pin), estimates prompt tokens
including the out-of-band carriers (top-level system, tools,
instructions), and when the decided tier provably cannot hold the prompt
moves the request to the lowest configured tier with a model whose
declared window fits, restricting the pick to fitting models when the
decided tier can keep it. Models with no resolvable window are never
escalated away from or onto, escalated decisions are never written as
session pins, and the decision records context_escalated plus the
original tier in spend logs.
Resolves LIT-6503
* fix(complexity_router): judge groups by smallest window, bound skips by bytes, filter adaptive picks
Review-round rework, one mechanism per finding. A group is judged by its
smallest resolvable deployment window, since the core router picks within
a group with no fit check. The counting skip is gated on UTF-8 byte
length, which BPE token counts can never exceed, so token-dense scripts
cannot slip past it; only a real tokenizer count ever moves a request and
a failed count leaves the placement alone. The fit facts now filter every
adaptive phase including cold start and the tier fallbacks. Window
questions adopt the declared provider and never resolve authenticating
providers, and a router instance without get_model_list degrades the gate
to a no-op. Tests rebuilt on real Router instances resolving deployment
model_info end to end, plus a full-path test through
async_get_available_deployment
UsersTable overrides DataTable's default noDataMessage with its own
EmptyState, so the row reads "No users found" rather than "No results".
Assert that, and pair it with the seeded user being absent so the check
cannot pass while the filter silently does nothing.
The batch start told the seed which LiteLLM_SpendLogs rows were its own, but
using it as a hard cutoff also dropped rows another pod had already persisted.
Those rows are only repaid by that pod's own increment, so if it died first the
window row stayed permanently under the recorded spend.
The seed now reads both sums in one scan and takes off this batch's own spend,
flooring at the pre-batch total for the case where its log rows have not landed
yet. Redis payloads keep an empty request_ids so a leader from before the field
was dropped can still merge what it pops during a rolling deploy.
Claude-Session: https://claude.ai/code/session_01QvQzYztinxj8ZuD5YxbVdL
Two assertions were checking the wrong thing. The anchoring tests read
getByRole("listbox"), which resolves to SelectPrimitive.List; that sits at
full content height inside the popup that clips and scrolls it, so the box
overlapped the trigger even when nothing visible did. Measure the popup.
The SSO-ID search expected zero rows, but DataTable renders a "No results"
message row when a filter matches nothing, so the count is one. Assert the
empty state the user actually sees.
GigaChat reports prompt_tokens and total_tokens after subtracting cached
tokens (the docs example is prompt_tokens=1, precached_prompt_tokens=37,
total_tokens=5, so the fields are disjoint, not a subset). Map to the
OpenAI convention by adding precached_prompt_tokens back onto prompt and
total while still surfacing it as prompt_tokens_details.cached_tokens.
The base added router_metadata to SpendLogsMetadata in #39001 without
updating this fixture, and its CI run never executed logging_testing,
so the job now fails on every branch merged with current staging.
The consolidated popup test only asserted the options never cover the
trigger, so opening above the trigger with room below it, the regression
PR #38554 fixed, would have passed. Split it back into a below-trigger
case and a cramped-viewport case. The header test accepted a single pixel
of vertical intersection; require the refresh control's centre to sit
within the tab row instead.
The migration smoke waited on `getByRole("button", { expanded: false })`
after clicking it. Playwright re-resolves that locator on every retry, so
once the clicked group flipped to expanded it matched the next collapsed
group instead, and the assertion could never pass. Count the remaining
collapsed groups and wait for that count to drop by one.
precached_prompt_tokens is a subset of prompt_tokens (OpenAI cached_tokens
semantics), so map it to prompt_tokens_details.cached_tokens instead of
adding it on top of prompt/total. Emit stream usage from any final chunk
carrying it rather than only finish_reason stop, which dropped tokens for
function_call and length streams. Merge auth metadata into a new dict in
the gigachat router handler instead of mutating the shared parsed-body
cache in place.
