Commit graph

32960 commits

Author SHA1 Message Date
yuneng-jiang
aa9f85545c
Merge pull request #21627 from BerriAI/litellm_service_key_vis
[Fix] Service Account Visibility for Team Members
2026-02-19 17:09:33 -08:00
yuneng-jiang
4475258426 address greptile review feedback (greploop iteration 1)
Refactor to fetch team objects once via _fetch_user_team_objects(),
then derive admin and member team IDs from the shared result. Eliminates
duplicate DB query between get_admin_team_ids and get_member_team_ids.
2026-02-19 16:53:48 -08:00
yuneng-jiang
c7fe400d4d [Fix] Service account visibility for team members
Regular team members could not see service accounts belonging to their
team. Additionally, former team members could still see service accounts
they created via the created_by filter after leaving the team.

- Add get_member_team_ids() to retrieve all teams a user belongs to
- Regular members now see team service accounts (user_id=NULL) but not
  other members' personal keys
- Scope created_by filter to the user's current team memberships
- Add 7 unit tests covering the new visibility rules
2026-02-19 16:29:43 -08:00
github-actions[bot]
2d39825868
chore: regenerate poetry.lock to match pyproject.toml (#21626)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2026-02-20 00:25:50 +00:00
jquinter
30a6b18c0d
Merge pull request #21625 from BerriAI/fix/regenerate-lock-drop-pat-approval
fix(ci): drop PAT_TOKEN_2 approval, use github.token for auto-merge
2026-02-19 21:20:02 -03:00
Julio Quinteros Pro
11d0fca0de fix(ci): drop PAT_TOKEN_2 approval step, use github.token for auto-merge
PAT_TOKEN_2 does not have the scope for addPullRequestReview.
github.token cannot approve its own PR either, so drop the approval
step entirely. Auto-merge with github.token is enough: the PR will
merge automatically once required CI checks pass.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 21:17:40 -03:00
jquinter
3a51d8bcb5
Merge pull request #21619 from BerriAI/fix/regenerate-lock-auto-approve
feat(ci): auto-approve and auto-merge the regenerated poetry.lock PR
2026-02-19 21:03:16 -03:00
Ishaan Jaffer
d4ae3c8370 ui new build 2026-02-19 15:57:19 -08:00
Ishaan Jaff
655973e8c8
feat(ui): show latency overhead for AI-suggested policy templates (#21620)
* feat(policy): add estimated_latency_ms to all policy templates

* feat(policy): add estimated_latency_ms to backup templates

* feat(ui): show latency overhead badge in AI policy suggestions
2026-02-19 15:45:07 -08:00
Julio Quinteros Pro
41776b0382 feat(ci): auto-approve and auto-merge the regenerated poetry.lock PR
Now that "Allow GitHub Actions to create and approve pull requests" is
enabled in repo settings:
- PR creation uses github.token (no secret needed)
- Approval uses PAT_TOKEN_2 (GitHub requires a different identity from
  the PR creator to approve)
- Auto-merge is enabled with --squash so the PR merges as soon as
  required checks pass

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 20:23:07 -03:00
jquinter
c32892dd46
Merge pull request #21618 from BerriAI/fix/regenerate-lock-pr-token
fix(ci): use PAT_TOKEN_2 for gh pr create in regenerate-lock workflow
2026-02-19 20:16:49 -03:00
Julio Quinteros Pro
9c70cd615a fix(ci): use PAT_TOKEN_2 for gh pr create
github.token cannot open PRs when "Allow GitHub Actions to create and
approve pull requests" is disabled in repo settings. PAT_TOKEN_2
bypasses that restriction.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 20:13:58 -03:00
jquinter
b88ecb11f0
Merge pull request #21617 from BerriAI/fix/regenerate-lock-no-update-flag
fix(ci): remove --no-update flag removed in Poetry 2.x
2026-02-19 20:09:57 -03:00
Julio Quinteros Pro
08b5907c9c fix(ci): remove --no-update flag removed in Poetry 2.x
The workflow fails with:
  The option "--no-update" does not exist

--no-update was removed in Poetry 2.x. Plain `poetry lock` is the
correct equivalent — it re-solves only what pyproject.toml requires
without upgrading already-locked packages.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 20:07:14 -03:00
Ishaan Jaff
3d8c042ca5
feat: prompt injection guardrail policy template (#21520)
* add semantic guard constants

* add SEMANTIC_GUARD enum value

* add keyword-based prompt injection policy template

* add semantic prompt injection route template

* add semantic guard route loader

* add semantic guard guardrail

* add semantic guard registration

* add semantic guard tests

* enhance SQL injection keyword category with more patterns and exceptions

* add standalone SQL injection policy template

* add SQL injection semantic guard route template

* add SQL injection guardrail tests

* remove standalone sql_injection policy template, use categories/prompt_injection_sql.yaml instead
2026-02-19 15:06:06 -08:00
jquinter
1edd6459b3
Merge pull request #21616 from BerriAI/fix/regenerate-lock-token-fallback
fix(ci): fall back to github.token when GH_TOKEN secret is not set
2026-02-19 20:01:46 -03:00
Julio Quinteros Pro
54470ec1d9 fix(ci): use github.token with explicit permissions instead of PAT secret
Drop the PAT_TOKEN_2 secret (whose scope is unknown) in favour of the
built-in github.token, which is always available. Grant it exactly the
two permissions it needs:
  - contents: write      → push the auto/regenerate-* branch
  - pull-requests: write → open the PR via gh cli

No external secret needed.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:56:59 -03:00
Julio Quinteros Pro
8cc50d6736 chore: fix stale GH_TOKEN comment 2026-02-19 19:55:44 -03:00
Julio Quinteros Pro
e590674083 fix(ci): use PAT_TOKEN_2 instead of non-existent GH_TOKEN secret
GH_TOKEN is not configured in this repository. The correct PAT secret
is PAT_TOKEN_2, which has the permissions needed to push branches and
open PRs.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:55:33 -03:00
Julio Quinteros Pro
755dd46d45 fix(ci): fall back to github.token when GH_TOKEN secret is not set
When secrets.GH_TOKEN is not configured, the workflow fails immediately with:
  "Input required and not supplied: token"

Using || github.token ensures a valid token is always available.
GH_TOKEN (PAT) is preferred when set; github.token is used as fallback.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:52:01 -03:00
jquinter
a6af0425ea
Merge pull request #21615 from BerriAI/fix/regenerate-poetry-lock-yaml-syntax
fix(ci): fix YAML syntax error in regenerate-poetry-lock workflow
2026-02-19 19:47:36 -03:00
Julio Quinteros Pro
8edc6effb1 fix(ci): fix YAML syntax error in regenerate-poetry-lock workflow
A heredoc inside \$() inside a double-quoted string inside a YAML multiline
run block breaks the YAML parser at line 60. Replaced with: write the PR
body to /tmp/pr-body.md using a standalone heredoc, then pass it via
gh pr create --body-file.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:45:14 -03:00
jquinter
f106c97bf2
Merge pull request #21609 from BerriAI/fix/delete-keys-return-failed-tokens
fix(key management): return failed_tokens in delete_verification_tokens response
2026-02-19 19:31:39 -03:00
jquinter
b32ba5902b
Merge pull request #21610 from BerriAI/fix/auto-regenerate-poetry-lock
ci: auto-regenerate poetry.lock when pyproject.toml changes on main
2026-02-19 19:29:08 -03:00
Julio Quinteros Pro
b9f36645f3 fix(key management): populate failed_tokens for admin partial-delete path
The admin bulk-delete path (prisma_client.delete_data) was never comparing
the DB return value against the requested tokens, so failed_tokens was
always empty for admins even when the DB silently skipped some tokens.

Adds the same mismatch check as the non-admin path, and a new test that
exercises admin bulk-delete returning fewer tokens than requested.

Fixes gap identified by Greptile review on #21609.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:28:03 -03:00
Julio Quinteros Pro
b9e79cc07b fix(ci): restrict GITHUB_TOKEN to contents:read via explicit permissions block
GitHub Advanced Security flagged that the workflow had no permissions block,
leaving GITHUB_TOKEN with its default broad scope. All write operations
(git push, gh pr create) already use GH_TOKEN (PAT), so the implicit
GITHUB_TOKEN only needs read access.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:24:57 -03:00
Julio Quinteros Pro
5681229e7c fix(ci): force-push bot branch to handle pre-existing branch from prior run
A re-run within the same second (or a leftover branch) would cause
`git push` to fail. Adding -f is safe since this is a bot-owned branch
that is immediately turned into a PR and never used for anything else.

Fixes inline suggestion from Greptile review.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:20:01 -03:00
Julio Quinteros Pro
dfbf14e626 fix(ci): pass GH_TOKEN to checkout so git push can create the branch
Without the token in the checkout step the subsequent `git push` uses the
default GITHUB_TOKEN which lacks permission to push new branches, causing
the workflow to fail silently. Fixes issue flagged by Greptile review.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:18:18 -03:00
Julio Quinteros Pro
d7d651dfeb ci: auto-regenerate poetry.lock when pyproject.toml changes on main
Adds a workflow that triggers whenever pyproject.toml is merged into main
and opens a PR with the refreshed lock file, fixing the recurring CI failure:
"pyproject.toml changed significantly since poetry.lock was last generated."

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:13:54 -03:00
Julio Quinteros Pro
079bc364bb fix(key management): return failed_tokens in delete_verification_tokens response
Instead of raising an exception when some requested tokens are not found
in the DB, populate `failed_tokens` in the return dict so callers can
inspect which keys were not deleted without catching a generic Exception.

Supersedes #12577.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-02-19 19:10:55 -03:00
Ishaan Jaff
c9cdce96fa
feat(policy): test playground for AI policy suggestions (#21608)
* fix aviation safety topic filter: remove overly broad exceptions, add cockpit access block words

* fix airline brand protection filter: add identifier words, competitor/ops block words, tighten exceptions

* feat(policy): add POST /policy/templates/test endpoint for testing guardrails before creating them

* feat(ui): add testPolicyTemplate networking function

* feat(ui): add test playground to AI policy suggestion modal

* test(policy): add tests for POST /policy/templates/test endpoint
2026-02-19 14:09:20 -08:00
Ishaan Jaff
779c355f1e
feat: add airline off-topic restriction policy template (#21607)
* fix aviation safety topic filter: remove overly broad exceptions, add cockpit access block words

* fix airline brand protection filter: add identifier words, competitor/ops block words, tighten exceptions

* feat: add airline off-topic restriction policy template

* feat: register airline off-topic restriction in policy templates backup

* test: add tests for airline off-topic restriction policy template

* fix: remove overly broad block words (explain, recommend, suggest, best) to reduce false positives

* test: add false-positive tests for recommend/explain airline queries
2026-02-19 13:45:59 -08:00
Ishaan Jaff
18f8a2cee3
docs: add latency overhead troubleshooting guide (#21603)
* add latency overhead troubleshooting doc

* add latency_overhead to troubleshooting sidebar

* docs: add x-litellm-overhead-duration-ms to latency troubleshooting guide
2026-02-19 12:42:33 -08:00
Ishaan Jaff
2c8fcf854a
docs: add latency overhead troubleshooting guide (#21600)
* add latency overhead troubleshooting doc

* add latency_overhead to troubleshooting sidebar
2026-02-19 12:34:23 -08:00
Ishaan Jaff
cd95e54c10
Add OpenAPI-to-MCP support via API and UI (#21575)
* add spec_path column to LiteLLM_MCPServerTable schema

* add spec_path to MCP request types and table model

* wire spec_path through build_mcp_server_from_table

* add openapi transport type constant

* add OpenAPI Spec as first-class transport option in create form

* add OpenAPI transport support to edit form with auto-detection

* support spec_path in connection status component

* support spec_path in tool configuration component

* support OpenAPI transport in test connection hook

* register OpenAPI tools on server add/update/reload

* preview OpenAPI tools in test/tools/list endpoint
2026-02-19 12:23:24 -08:00
Ishaan Jaff
e9a07347dc
fix: reduce proxy overhead for large base64 payloads (#21594)
* fix aviation safety topic filter: remove overly broad exceptions, add cockpit access block words

* fix airline brand protection filter: add identifier words, competitor/ops block words, tighten exceptions

* add constants for large payload handling and detailed timing

* add base64 truncation for logging payloads

* use shallow copy for messages, track copy and callback timing

* add callback duration and detailed timing to response metadata

* add callback duration header, size-gate debug logging, detailed timing headers

* add tests for callback timing, base64 truncation, and detailed timing

* fix code quality: extract helpers, fix regex, clean up imports

* rewrite _truncate_base64_in_value iteratively to satisfy recursive detector
2026-02-19 12:10:52 -08:00
Ishaan Jaff
b209b11522
feat: AI policy template suggestions (#21589)
* fix aviation safety topic filter: remove overly broad exceptions, add cockpit access block words

* fix airline brand protection filter: add identifier words, competitor/ops block words, tighten exceptions

* add example_sentences to all policy templates + topic-filtering and prompt-injection templates

* add policy_endpoints package with AI policy suggester

* update test patch targets for policy_endpoints package move

* add unit tests for AI policy suggester

* add suggestPolicyTemplates networking function

* add AI suggestion modal component

* add Use AI button and template loading callback to PolicyTemplates

* wire up AI suggestion modal in policies page

* fix policy_templates_backup.json path after package move

* add estimated_latency field to all policy templates

* use llm_router and accept model parameter in ai_policy_suggester

* add model param to suggest templates endpoint

* pass model param in suggestPolicyTemplates

* polish ai suggestion modal: model selector, auto-growing textareas, latency badges

* add template queue for processing multiple AI-suggested templates

* show template progress badge in guardrail selection modal
2026-02-19 12:00:26 -08:00
michelligabriele
16cfdccc7b
fix(bedrock): add Accept header for AgentCore MCP server requests (#21551)
AgentCore MCP server endpoints require the Accept header to contain
both application/json and text/event-stream per the MCP specification
(Streamable HTTP transport). Without this header, requests are rejected
with a 406 Not Acceptable error (JSON-RPC code -32011).

Sets the Accept header at the top of sign_request() so both JWT/Bearer
and SigV4 authentication paths include it.
2026-02-19 11:35:14 -08:00
milan-berri
506d1dff43
fix: handle deprovisioning operations without path field (#21571)
* fix(scim): handle deprovisioning operations without path field

When SCIM providers send deprovisioning requests without a path field
(e.g., {"op": "replace", "value": {"active": false}}), the code was
storing the value under an empty string key in metadata.

This fix:
- Detects operations with no path where value is a dict
- Extracts and handles known fields like 'active' correctly
- Sets metadata["scim_active"] = false instead of metadata[""] = {"active": false}

Fixes: SCIM deprovisioning creating empty string keys in user metadata

* fix(scim): handle all known fields in operations without path

Extended the fix to handle all SCIM fields (not just active) when
operations have no path field:
- active -> scim_active
- displayName -> user_alias
- externalId -> user_id
- name.givenName/familyName -> scim_metadata

Added comprehensive test for multiple fields without path.

Addresses Greptile review feedback on RFC 7644 compliance.

* trigger PR update
2026-02-19 11:09:04 -08:00
Sameer Kankute
ddc1371cea
Merge pull request #21588 from BerriAI/geminii_blog
Fix release
2026-02-20 00:31:01 +05:30
Sameer Kankute
4d392cacb8 Fix release 2026-02-20 00:27:12 +05:30
yuneng-jiang
aa4a95c889
Merge pull request #21587 from BerriAI/litellm_fix_key_delete_feb19
[Infra] Add project_id to DeletedVerificationTable
2026-02-19 10:51:28 -08:00
yuneng-jiang
6c0cc4fb4f adding build 2026-02-19 10:50:47 -08:00
yuneng-jiang
2eded1dda6 bump: version 0.4.43 → 0.4.44 2026-02-19 10:50:23 -08:00
yuneng-jiang
a5fba84ae3
Merge pull request #21586 from BerriAI/infra_feb19
[Infra] Fixing Merge Artifacts
2026-02-19 10:47:39 -08:00
yuneng-jiang
2d928f8c54 fixing merge 2026-02-19 10:46:58 -08:00
yuneng-jiang
bac1b6b2e0
Merge pull request #21545 from BerriAI/litellm_key_last_active_tracking
[Feature] Key Last Active Tracking
2026-02-19 10:29:23 -08:00
yuneng-jiang
caa931b167 adding build 2026-02-19 10:29:01 -08:00
yuneng-jiang
9eaaa9740b bump: version 0.4.42 → 0.4.43 2026-02-19 10:28:35 -08:00
yuneng-jiang
c911cfbabf Merge remote-tracking branch 'origin' into litellm_key_last_active_tracking 2026-02-19 10:27:48 -08:00