scriptoriumadmin
  • Joined on 2026-03-01
scriptoriumadmin synced commits to refs/pull/38304/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
74bc1efe65 Merge branch 'litellm_internal_staging' into litellm_fix_staging_ci_regressions
02c1c45b53 fix(complexity_router): route client housekeeping calls to the cheapest tier (#38598)
b3322e30a7 fix(anthropic): handle per-level reasoning_effort flags without supports_reasoning (#38618)
39e5b0c2d1 fix(anthropic): drop and self-heal empty thinking blocks on /v1/messages (#38625)
bd7e9c1997 Merge pull request #38624 from BerriAI/litellm_/circleci-regressions-review-f0a542
Compare 534 commits »
scriptoriumadmin synced commits to refs/pull/38319/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
f814945d4c Merge pull request #38917 from BerriAI/litellm_pin_wolfi_python313
b518be45fb Merge pull request #38997 from BerriAI/litellm_add_responses_input_tokens_endpoint
40edeaaecb fix(otel): emit cache token counts on OTel v2 LLM spans (#38716)
b7da471784 fix(count_tokens): price an inline file block instead of raising on it
Compare 638 commits »
scriptoriumadmin synced commits to refs/pull/38324/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
62f032cca5 fix(proxy): keep passthrough logging metadata and model_info dicts when team callbacks are wired (#39216)
364c9be2e0 Merge pull request #39211 from BerriAI/litellm_fix_default_api_key_limit_leak
4b87fd5718 fix: normalize provider-specific cache token fields in OTel v2 usage (#39202)
2b616fc479 feat(scim): add placeholder listing and merge so a shadowed account can be healed (#39231)
Compare 1122 commits »
scriptoriumadmin synced commits to refs/pull/38333/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
91789da296 fix(vertex_ai): verify embedded thought signature checksum
ae79d98250 style(vertex_ai): remove redundant signature comments
f3e0db6859 fix(vertex_ai): satisfy TRY300 gate and pin extractor scope
b80f305430 fix(vertex_ai): reject client-mangled thought_signature embedded in tool_call_id
4ba8517134 Merge pull request #38722 from BerriAI/litellm_bedrock_guardrail_stream_audit
Compare 956 commits »
scriptoriumadmin synced commits to refs/pull/38333/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
9c05c158cb Merge pull request #39518 from BerriAI/litellm_techdebt_20260903
3d00ad3f29 Merge pull request #39847 from BerriAI/litellm_lit_5730_bedrock_batch_cancel_e2e
8b6ea72845 feat(shadow_eval): scope a job to model groups, ANDed with its key, team, and user targets (#39828)
e7dd524a3c feat(otel): stamp litellm.request.route on the LLM call span (#39698)
Compare 2007 commits »
scriptoriumadmin synced commits to refs/pull/38334/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
8699998c9e Merge pull request #39625 from BerriAI/litellm_lit6873_vertex_passthrough_api_version
1ea95fa18f Merge pull request #39617 from BerriAI/litellm_fix_responses_background_stream_flake
ed9865c945 Merge pull request #39604 from BerriAI/litellm_lit_4738_users_search_by_id
5a821b593c Merge pull request #38440 from BerriAI/litellm_prompt_registry_env
Compare 1600 commits »
scriptoriumadmin synced commits to refs/pull/38336/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:02 +00:00
2ce4e3f8a9 fix(guardrails): run apply_guardrail-only providers in logging_only mode (#39297)
8588a2ea42 fix(docker): install saml extra in litellm-backend image (#39291)
31ca4ddf32 Merge pull request #39166 from BerriAI/litellm_bedrock_bearer_token_converse_crash
e2c3f51c46 fix(search): forward search-tool params through the router, complete Parallel AI v1 param mapping (#37883)
Compare 1144 commits »
scriptoriumadmin synced commits to refs/pull/38284/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
c8635ecc67 feat: page the public model hub table off /public/v1/model_hub, keeping every filter (#39691)
eb67e5402b test(ocr): complete Rust unit test parity (#39689)
ee08c36fc0 refactor(tests): restructure rust python harness around strategy definitions (#39628)
8fc0663198 fix(ci): pin setup-uv to v10.0.1 (#39111)
Compare 1733 commits »
scriptoriumadmin synced commits to refs/pull/38285/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
9e01bd1441 fix(azure): send the deployment name as the body model on v1 image routes
scriptoriumadmin synced commits to refs/pull/38278/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
d23bec84c4 Merge pull request #39196 from BerriAI/litellm_guardrail_usage_cost_rollup
6dff3a5f72 fix(complexity_router): fall back to a live peer when the decided tier model is fully cooled down (#39675)
df68edca76 Merge pull request #39808 from BerriAI/litellm_lit_5379_jwt_mapping_cache_invalidation
922659fb15 Merge pull request #39388 from BerriAI/litellm_registry_audit_2026_09_02
Compare 1906 commits »
scriptoriumadmin synced commits to refs/pull/38288/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
59d42d36e6 fix(headroom): bound the /v1/compress and /v1/retrieve calls with a timeout (#39527)
be76dfad9c Merge pull request #35853 from BerriAI/litellm_retry_policy_503
6960a42008 Merge pull request #36575 from BerriAI/litellm_mcp_stateless_follow_up_zdr
aa3d59d086 fix(anthropic): never carry cache_control on translated thinking blocks (#39815)
Compare 1779 commits »
scriptoriumadmin synced commits to refs/pull/38289/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
3d4ba87e8c fix(proxy): guard extract_key_hash/extract_key_alias against a non-Mapping metadata field
41be4a8782 fix(proxy): require an unforgeable marker to trust litellm_metadata
9813335651 chore: retrigger CI (zizmor cancelled by infra/concurrency at the queue stage, not a real failure)
890c4a43d5 fix(proxy): stop a caller-supplied tag from shadowing a policy-backed identity tag
effe9ffa6f chore: retrigger CI (previous run's jobs were cancelled by infra/concurrency, not a real failure)
Compare 1310 commits »
scriptoriumadmin synced commits to refs/pull/38289/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
d23bec84c4 Merge pull request #39196 from BerriAI/litellm_guardrail_usage_cost_rollup
6dff3a5f72 fix(complexity_router): fall back to a live peer when the decided tier model is fully cooled down (#39675)
df68edca76 Merge pull request #39808 from BerriAI/litellm_lit_5379_jwt_mapping_cache_invalidation
922659fb15 Merge pull request #39388 from BerriAI/litellm_registry_audit_2026_09_02
Compare 1802 commits »
scriptoriumadmin synced commits to refs/pull/38290/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
11a02b9581 Merge pull request #36260 from eeshsaxena/fix-credential-endpoints-raise-not-return
f83b6029e3 Merge branch 'litellm_internal_staging' into fix-credential-endpoints-raise-not-return
418d0e79ba fix(proxy): answer 404 when deleting a credential that was never stored
658f50663d fix(ui): keep Virtual Keys list state in the URL so it survives leaving the page (#39481)
Compare 1474 commits »
scriptoriumadmin synced commits to refs/pull/38292/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:01 +00:00
44a1ac32b6 fix(proxy): reserve one concurrency unit per genuinely concurrent batch branch, not one per dispatch
c817e0150c fix(proxy): reserve exactly one concurrency unit per abatch_completion dispatch, not one per branch
f825c03b67 fix(proxy): don't release a live sibling's slot when this hop's own context resolves to nothing configured, and lock in multi-policy release
180b5fd69b fix(proxy): stop the terminal-release fallback and a shared admission snapshot from crossing abatch_completion sibling branches
780def9194 fix(proxy): release exactly the concurrency reservation a completing hop's own admission made, not every pending entry
Compare 1423 commits »
scriptoriumadmin synced commits to refs/pull/38255/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:00 +00:00
b518be45fb Merge pull request #38997 from BerriAI/litellm_add_responses_input_tokens_endpoint
40edeaaecb fix(otel): emit cache token counts on OTel v2 LLM spans (#38716)
b7da471784 fix(count_tokens): price an inline file block instead of raising on it
be3386f3ce Merge pull request #38995 from BerriAI/litellm_openai_wif
Compare 636 commits »
scriptoriumadmin synced commits to refs/pull/38256/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:00 +00:00
78ff5ac9cd feat(router): arm safeguard-refusal fallback on generic chains when no content-policy list exists (#39274)
8441dd6e8c fix(proxy): keep SpendLogs and callback session ids in sync when the request has none (#39450)
92edcb90db fix: keep litellm importable on Python 3.10 and guard 3.11-only typing imports in CI (#39448)
4f7b20ec10 fix(guardrails): skip streaming guardrail rounds that re-scan cleared output (#39386)
Compare 1426 commits »
scriptoriumadmin synced commits to refs/pull/38259/head at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:00 +00:00
0568cf094a address post-merge check failures
7548cea3ee Merge remote-tracking branch 'upstream/litellm_internal_staging' into fix/vector-store-list-preserves-config-sourced
d4f6b4491d Merge pull request #38656 from aaaaaandrew/litellm_preserve_stream_selected_model
18830667b2 Merge pull request #38691 from BerriAI/litellm_messages_cost_alias_pricing
1e86532c3d feat(auto_router): write and preview the classifier prompt an edited tier set sends (#38605)
Compare 956 commits »
scriptoriumadmin synced commits to refs/pull/38260/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:00 +00:00
40edeaaecb fix(otel): emit cache token counts on OTel v2 LLM spans (#38716)
be3386f3ce Merge pull request #38995 from BerriAI/litellm_openai_wif
66295e7da7 Merge pull request #34696 from cat0825/fix/34379-unblock-customer
1249f84b10 fix(vertex_ai): graft default vertex path when api_base has a version-only path (#38986)
Compare 629 commits »
scriptoriumadmin synced commits to refs/pull/38245/merge at scriptoriumadmin/litellm from mirror 2026-09-05 08:10:00 +00:00
9153bc6236 fix(openai): do not pass ssl_verify to extra_body and apply per-call TLS config to HTTP client
d23bec84c4 Merge pull request #39196 from BerriAI/litellm_guardrail_usage_cost_rollup
6dff3a5f72 fix(complexity_router): fall back to a live peer when the decided tier model is fully cooled down (#39675)
df68edca76 Merge pull request #39808 from BerriAI/litellm_lit_5379_jwt_mapping_cache_invalidation
Compare 1924 commits »