- New troubleshoot page and blog post with step-by-step comparison workflow
- Screenshots under static/img/cost-discrepancy-debug
- Link from spend tracking; sidebar entry under Troubleshooting
- Flowchart SVG: Path B connectors below box; clarify LiteLLM schedules customer calls when stuck
Made-with: Cursor
Brings the date-range branch in line with the non-date-range branch which
already hashes sk- prefixed tokens before querying. Adds coverage for
filter-combination behavior in view_spend_logs.
- Add LITELLM_OIDC_ALLOWED_CREDENTIAL_DIRS to the environment variables
reference so the documentation test passes.
- Annotate the values variable in _reject_os_environ_references so it
accepts both dict.values() and list iterables.
- Log a warning when dropping callback params that carry os.environ/
references so operators notice the misconfiguration.
- Require absolute paths in oidc/file/ and correct the documented
example to use the leading-slash form.
- Drop the unused return value from _reject_os_environ_references.
- Reject os.environ/ references supplied via /health/test_connection
request params instead of resolving them; config-sourced values are
already resolved before reaching the endpoint.
- Skip os.environ/ references in dynamic callback params loaded from
per-request metadata.
- Constrain oidc/file/ to an allowed credential directory allowlist
(defaults to /var/run/secrets and /run/secrets, overridable via
LITELLM_OIDC_ALLOWED_CREDENTIAL_DIRS).
Unit Tests: Proxy DB Operations / proxy-db (auth-checks, tests/proxy_unit_tests/test_auth_checks.py tests/proxy_unit_tests/test_user_api_key_auth.py, 20, 8) (push) Waiting to run
Unit Tests: Proxy DB Operations / proxy-db (remaining, tests/proxy_unit_tests --ignore=tests/proxy_unit_tests/test_key_generate_prisma.py --ignore=tests/proxy_unit_tests/test_auth_checks.py --ignore=tests/proxy_unit_tests/test_user_api_key_auth.py, 20, 8) (push) Waiting to run
Budget table entries (team members, end-users) used duration_in_seconds()
for a sliding-window reset, while keys/users/teams used calendar-aligned
get_budget_reset_time(). This made "30d" and "1mo" mean different things
depending on entity type. Now both paths use get_budget_reset_time() for
consistent calendar-aligned resets (e.g. "30d" → 1st of next month).
Fixes#25432
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The model_max_budget limiter tracks spend in one code path
(async_log_success_event) and enforces budget limits in another
(is_key_within_model_budget via user_api_key_auth). These two paths
used different model name formats to build cache keys:
- Tracking used standard_logging_payload["model"], which is the
deployment-level model name (e.g. "vertex_ai/claude-opus-4-6@default")
- Enforcement used request_data["model"], which is the model group
alias (e.g. "claude-opus-4-6")
Because the cache keys never matched, the enforcement path always read
None for current spend, silently allowing all requests through even
after the budget was exceeded. This affected any provider that decorates
model names with provider prefixes or version suffixes (Vertex AI,
Bedrock, etc.).
Fix: use model_group (the user-facing alias) from StandardLoggingPayload
for spend tracking, falling back to model when model_group is None.
This aligns the tracking cache key with the enforcement cache key.
Fixes the same root cause reported in #15223 and #10052.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two code paths in key_management_endpoints.py call hash_token()
unconditionally when invalidating the user_api_key_cache after a key
update. When the caller passes a pre-hashed token ID (not an sk-
prefixed key), hash_token() double-hashes it, producing a cache key
that does not match the actual cached entry. Cache invalidation
silently fails.
This is compounded by update_cache() which writes the stale cached key
object back with a fresh 60s TTL after every successful request,
preventing natural TTL expiry. The stale entry (with outdated fields
like max_budget=None) persists indefinitely under load.
PR #24969 fixed this in update_key_fn but missed two other call sites:
- _process_single_key_update (bulk update path)
- _execute_virtual_key_regeneration (key rotation path)
Fix: replace hash_token() with _hash_token_if_needed() in both
locations, matching the pattern already used elsewhere in the file.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When the advisor loop hits max_uses, inject a tool_result error so the executor
sees the cap and continues without further advice — matches Anthropic server-side
behaviour (error_code: max_uses_exceeded).
Unit Tests: Proxy DB Operations / proxy-db (auth-checks, tests/proxy_unit_tests/test_auth_checks.py tests/proxy_unit_tests/test_user_api_key_auth.py, 20, 8) (push) Has been cancelled
Unit Tests: Proxy DB Operations / proxy-db (remaining, tests/proxy_unit_tests --ignore=tests/proxy_unit_tests/test_key_generate_prisma.py --ignore=tests/proxy_unit_tests/test_auth_checks.py --ignore=tests/proxy_unit_tests/test_user_api_key_auth.py, 20, 8) (push) Has been cancelled
The new effectiveGuardrails computation derived its element type from
`info.metadata?.guardrails`, which is implicitly typed `any[]`, so the
downstream `.map((name) => ...)` callback inferred `name` as implicit
`any` and broke the production typecheck.
Renames the new per-guardrail opt-out field from `disabled_global_guardrails`
to `opted_out_global_guardrails` to eliminate the one-character collision with
the legacy `disable_global_guardrails` boolean kill switch. Adds a type guard
on the new gate so a misnamed bool can't crash the guardrail check. Filters
duplicates out of the team-edit guardrail display for legacy teams that have a
global name persisted in `metadata.guardrails` from before this PR. Drops the
unused `isGuardrailsLoading` and `guardrailsError` destructures left in
AddModelForm after the hook refactor.
Adds Python tests for the new gate behavior (root, litellm_metadata, metadata,
non-matching name, empty list, malformed bool value, opt-in coexistence) and
extends useGuardrails.test.ts to exercise the global / optional partition
logic that the rebuilt hook performs in its `select` transform.
Wires the legacy kill switch and the new opt-out list together in the team
edit form so they can never fall out of sync:
- Toggling the kill switch reactively updates the Guardrails Select via
`onValuesChange` — switch on strips all globals from the selection, switch
off re-adds them. Existing opt-in extras are preserved either way.
- When the switch is on, global options in the Select are individually
disabled (greyed out) so the user can still manage opt-in guardrails but
cannot accidentally re-enable a global the kill switch is bypassing.
- The save handler writes both fields together: `disable_global_guardrails`
reflects the switch, and `opted_out_global_guardrails` is set to either
every global (when the switch is on) or the user's explicit opt-outs.
- `effectiveGuardrails` for the form's initialValues honors the kill switch
on legacy teams so the form opens in a state that matches what the runtime
gate is actually doing — fixes the visual lie where chips appeared active
while the switch was bypassing them.
The backend gate already reads the list as the primary path with the bool
as a fallback, so untouched legacy teams keep working until they get edited,
at which point they migrate naturally.
The team edit form computed its `guardrails` initialValues from
`globalGuardrailNames`, which is empty until `useGuardrails` resolves.
Because Ant Design's `<Form initialValues>` is read once on mount, a
user who clicked Edit before the query resolved would see globals
missing from the Select and could save that stale state, silently
opting the team out of every global guardrail.
Gate the form on `!isGuardrailsLoading` so initialValues always sees
the resolved set.
Replace the team info page's tags-mode guardrails Select with a grouped
multi-select that splits guardrails into "Global" (default_on=true) and
"Other" sections. On submit, derive both metadata fields from the single
selection array:
- metadata.guardrails: non-global opt-ins (additive)
- metadata.disabled_global_guardrails: globals the team has opted out of
The legacy disable_global_guardrails kill-switch toggle stays alongside
the multiselect and remains semantically distinct: it kills all global
guardrails including any added in the future, while the new list lets
new globals auto-apply unless explicitly excluded.
Other changes:
- Switch from a manual useEffect/useState guardrails fetch to the
consolidated useGuardrails hook
- Replace stale "Select existing guardrails or enter new ones" help
text with a single tooltip on each section's info icon
- Render selected chips with green/blue color coding via tagRender so
global vs non-global is visible without opening the dropdown
- Drop the redundant [Global] tag from inside the OptGroup options
(the group label already conveys it)
- Update the read-only display to show the effective set (globals not
opted out + additive opt-ins) with [Global] suffix on globals
Extend useGuardrails to return the full guardrail objects plus derived
globalGuardrailNames / optionalGuardrailNames sets via React Query's
select option, instead of just an array of names. Update its existing
consumer (AddModelForm) to extract names from the new shape.
The previous shape was tailored to AddModelForm's single use case
(populate a Select with names). The team info per-guardrail opt-out
work needs default_on per guardrail to split globals from non-globals,
which the old shape couldn't provide. Consolidating into the existing
hook gives both consumers one source of truth and one React Query
cache entry instead of two parallel fetches.
- useGuardrails.ts: rewrite return type, derive global/optional sets
in select(); preserve the existing query key and auth-gate semantics
- AddModelForm.tsx: extract names from data?.guardrails.map(...)
- AddModelForm.test.tsx: update mock to return the new shape (also
fixes a pre-existing shape mismatch in the mock)
- useGuardrails.test.ts: update 3 assertions to read names via
data?.guardrails.map(...) instead of asserting against the flat array
Add disabled_global_guardrails list field to team metadata that
selectively skips named globals at request time. Coexists with the
existing disable_global_guardrails boolean kill switch — the boolean
kills all globals (including future ones), the new list selectively
skips named ones (new globals auto-apply).
The field lives in the team metadata JSON column; no schema migration.
- litellm/proxy/litellm_pre_call_utils.py: propagate the field from
team_metadata into per-request data.metadata
- litellm/integrations/custom_guardrail.py: new
get_disabled_global_guardrails_from_metadata helper plus a
scoped-to-globals early return at the top of should_run_guardrail
