Commit graph

4637 commits

Author SHA1 Message Date
Michael Riad Zaky
4e26835098 Reorder counter invalidation to run after DB write 2026-04-30 17:50:58 -07:00
Michael Riad Zaky
fed5f36a3d Invalidate spend counters on budget reset 2026-04-30 17:50:58 -07:00
Michael Riad Zaky
9f08db91f9 Refresh Redis TTL on counter writes and skip stale in-memory on Redis miss 2026-04-30 17:50:58 -07:00
user
694fadd175 fix budget reservation review findings 2026-04-30 17:38:18 -07:00
user
2c852ba2b1 fix(sso): tighten oauth_state cookie — Secure flag + PKCE-only set
Two Greptile review findings addressed:

1. (P1, security) The ``litellm_oauth_state`` cookie is the sole
   guard against Login-CSRF in the PKCE flow but was set without the
   ``Secure`` attribute, so a network observer on plain HTTP could
   read and replay it — bypassing the protection this PR adds.

   Thread the originating ``Request`` down through
   ``get_sso_login_redirect`` and ``get_generic_sso_redirect_response``
   and set ``Secure`` based on ``request.url.scheme == "https"``.
   When no request is supplied (programmatic callers / tests) default
   to ``Secure=True`` — production-safe.  Local HTTP dev still works
   because the request scheme is observed at runtime.

2. (P2) The cookie was set unconditionally, but the callback only
   validates it inside the PKCE branch.  Two concurrent SSO sessions
   (one PKCE, one plain) could overwrite each other's state cookie
   and produce spurious 400s for the plain-flow user.

   Move the ``set_cookie`` call inside the existing
   ``if code_verifier and "state" in redirect_params`` block so the
   cookie is only written when PKCE is active and the validation
   will actually fire.

Tests cover both paths: PKCE-on (cookie set with Secure default),
PKCE-off (cookie not set), and HTTP dev request (Secure dropped so
the browser will actually attach the cookie on the callback hop).
2026-05-01 00:36:51 +00:00
harish-berri
8671ec636b fix import error 2026-05-01 00:29:13 +00:00
user
06502d19a7 test(vector stores): allow primitive rag depth boundary 2026-04-30 17:28:02 -07:00
harish-berri
7c8fe86fd9
Merge branch 'litellm_internal_staging' into litellm_token_verification_query_opt 2026-04-30 17:25:12 -07:00
user
00442e653c chore(sso): bind generic SSO state to a session cookie
The Generic SSO PKCE flow used the URL ``state`` parameter as the
cache key for the PKCE ``code_verifier`` without binding the state
to the caller's browser.  An attacker who pre-minted a state and
cached a verifier under it could hand the resulting login link to a
victim; the victim's auth code would then be exchanged with the
attacker's verifier on the callback, producing an access token
under the attacker's control (Login CSRF / token theft).

The non-PKCE branch is unaffected because it delegates to
fastapi-sso's ``verify_and_process``, which performs its own
session-cookie check.  The PKCE branch bypasses that helper, which
is exactly the gap this commit closes.

Two-part fix in ``ui_sso.py``:

- ``get_generic_sso_redirect_response`` now sets a
  ``litellm_oauth_state`` cookie (HttpOnly, SameSite=Lax, 10-min TTL)
  carrying the state value used in the redirect URL.  The cookie is
  set on the redirect response just like the existing
  ``litellm_cp_return_to`` cookie a few lines earlier in the file.
- ``get_generic_sso_response`` validates ``request.cookies.get(
  "litellm_oauth_state")`` against ``request.query_params.get(
  "state")`` via ``secrets.compare_digest`` before invoking the
  PKCE token exchange.  Mismatch (or either being missing) raises a
  ``ProxyException`` with HTTP 400.

The pre-existing TODO above the redirect logic ("state should be a
random string and added to the user session with cookie") is now
addressed and removed.

Tests cover the redirect-side cookie set, the missing-cookie reject
shape, the URL/cookie-mismatch reject shape, and the matching-cookie
happy path.
2026-05-01 00:19:13 +00:00
user
9db8ecac12 update budget reservation auth test expectation 2026-04-30 17:13:49 -07:00
yuneng-jiang
bdcc23853c
Merge pull request #26835 from stuxf/codex/cli-sso-flow-binding
chore(cli): tighten CLI SSO session flow
2026-04-30 17:10:27 -07:00
yuneng-jiang
15b7386859
Merge pull request #26815 from stuxf/fix/get-image-lfi-ssrf
chore(proxy): contain UI_LOGO_PATH / LITELLM_FAVICON_URL on unauthenticated asset endpoints
2026-04-30 17:10:15 -07:00
user
fce86d1334 fix budget reservation greptile findings 2026-04-30 17:08:45 -07:00
yuneng-jiang
71d5015975
Merge pull request #26827 from stuxf/fix/passthrough-auth-default
chore(passthrough): default auth=True and drop enterprise gate on the safe option
2026-04-30 17:06:37 -07:00
user
2922da9b64 test(vector stores): cover azure passthrough guard 2026-04-30 17:00:43 -07:00
user
f8d187785d finalize invalidated budget reservations 2026-04-30 16:52:09 -07:00
Michael Riad Zaky
053e040171 run pre_call_hook on Google generateContent endpoints 2026-04-30 16:43:42 -07:00
user
32272908d3 test(vector stores): isolate provider-native guard case 2026-04-30 16:41:13 -07:00
user
49ccb3369c test(vector stores): pin rag scan depth boundary 2026-04-30 16:33:42 -07:00
user
e034935b53 fix budget reservation window fallback races 2026-04-30 16:26:43 -07:00
user
1201a0ba5c test(vector stores): pin no-db registry fallback case 2026-04-30 16:16:37 -07:00
Michael Riad Zaky
47b2832d6f test: replace subprocess startup-import diff with static source scan 2026-04-30 16:15:46 -07:00
user
1373ae1021 fix budget tag spend counter reconciliation 2026-04-30 16:09:38 -07:00
user
ce0c557012 chore(vector stores): address access review followups 2026-04-30 16:09:26 -07:00
user
363c0de6f7 chore(vector stores): address tenant guard followups 2026-04-30 15:13:24 -07:00
user
d3ab59e059 chore(vector stores): tighten managed store access 2026-04-30 15:04:25 -07:00
user
96a283ed0f guard reservation invalidation cleanup 2026-04-30 15:00:35 -07:00
michelligabriele
0f98f3754f
fix(proxy): reject user_id=None on non-admin analytics endpoints (cross-tenant disclosure) 2026-04-30 23:54:01 +02:00
user
8311456cfc invalidate reservations after release cleanup failure 2026-04-30 14:48:55 -07:00
user
3800596d08 fix(proxy): stabilize lazy openapi snapshot ids 2026-04-30 14:42:50 -07:00
user
719e891c3a coalesce malformed window reservation seeding 2026-04-30 14:38:37 -07:00
user
15d4d51453 chore(callbacks): guard dynamic integration hosts 2026-04-30 14:27:19 -07:00
user
5521af096e preserve database failure during reservation cleanup 2026-04-30 14:23:57 -07:00
user
0b71282985 address budget reservation review findings 2026-04-30 14:06:42 -07:00
user
46183e6dc9 Merge remote-tracking branch 'origin/litellm_internal_staging' into codex/budget-race-enforcement-snapshot 2026-04-30 13:56:16 -07:00
user
0794ae67be avoid direct budget reservation db lookups 2026-04-30 13:49:59 -07:00
yuneng-jiang
256e05e474
Merge pull request #26849 from stuxf/fix/mcp-oauth-discovery-ssrf
chore(mcp): SSRF guard on OAuth metadata discovery follow-up fetches
2026-04-30 13:44:16 -07:00
yuneng-jiang
174c770b07
Merge pull request #26836 from stuxf/fix/byok-credential-encryption
chore(mcp): encrypt user-scoped MCP credentials at rest
2026-04-30 13:42:57 -07:00
user
ca50868b75 harden end-user and tag budget reservations 2026-04-30 13:37:10 -07:00
yuneng-jiang
4ff8f0e901
Merge pull request #26851 from stuxf/codex/fix-callback-env-secret-resolution
chore(proxy): block env callback refs in key metadata
2026-04-30 13:11:32 -07:00
yuneng-jiang
aa76ab2df7
Merge pull request #26862 from stuxf/codex/control-field-sanitization
chore(proxy): harden request control fields
2026-04-30 13:10:58 -07:00
Michael-RZ-Berri
9637d8c17b
Merge pull request #26802 from BerriAI/litellm_lazyLoadedFrontPage
[Feat / Fix] Lazy loaded imports, lazy loaded front page
2026-04-30 13:04:42 -07:00
user
51a3e90451 fix(mcp): reuse safe URL fetch for OAuth discovery 2026-04-30 12:42:52 -07:00
user
35bbca60b0 chore(proxy): default sensitive routes to auth 2026-04-30 12:22:48 -07:00
user
b31182f62a chore(proxy): preserve readiness compatibility 2026-04-30 12:05:09 -07:00
harish-berri
8df24b5413
Merge branch 'litellm_internal_staging' into litellm_token_verification_query_opt 2026-04-30 12:04:08 -07:00
yuneng-jiang
3c060364fb
Merge pull request #26840 from stuxf/codex/mcp-oauth-root-visibility
chore(mcp): tighten OAuth root endpoint resolution
2026-04-30 11:59:03 -07:00
yuneng-jiang
a9db887bdd
Merge pull request #26843 from stuxf/codex/fix-onboarding-invite-token
chore(auth): harden invite-link onboarding token flow
2026-04-30 11:56:26 -07:00
user
c8325ff007 chore(proxy): guard sensitive public endpoints 2026-04-30 11:52:47 -07:00
yuneng-jiang
0efa8b8828
Merge pull request #26854 from stuxf/fix/team-authz-available-team-bypass
chore(team): close authz bypass via the available-team check
2026-04-30 11:47:19 -07:00