Adds object_permission.skills to keys and teams, enforces it on
/claude-code/marketplace.json?key=, /claude-code/plugins and
/claude-code/plugins/{name}, and exposes an Allowed Skills selector in
the key and team create/edit forms of the Admin UI
Co-authored-by: yassin <yassin@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): reject initialize with 403 when the key grants no MCP servers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(mcp): e2e expects 403 initialize for a key with no MCP servers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): mention IP filtering in the no-servers initialize denial and keep zero-grant tool coverage
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: mateo <mateo@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(rust_bridge): count budget-check input tokens in Rust on all LLM routes
Rust counts input tokens from the raw JSON body with the GIL released inside the existing budget reservation, covering every LLM route the auth dependency guards. It only fires for models on the Anthropic tokenizer when a budget is set, and Python counts whenever Rust is off, missing, or declines a body shape.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* perf(rust): count byte-level BPE tokens without the GPT-2 split regex (#40594)
The oniguruma run of the ByteLevel pre-tokenizer regex is about 90% of
encode_fast on a 100k token body (100 ms of the ~110 ms Rust admission
count in the gateway pod). A hand-written scanner that yields the same
pieces, then feeds the model directly, counts the same text in 10 ms.
It only engages for tokenizers with the Anthropic shape (optional NFKC,
ByteLevel without prefix space, no post-processor) and falls back to the
full encoder when the text contains an added token. Parity with
encode_fast is tested on random texts, the pieces are compared with the
real pre-tokenizer, and the \p{L}/\p{N}/\s tables are checked against
oniguruma for every code point.
NFKC runs through unicode-normalization-alignments, the crate and
Unicode tables NormalizedString::nfkc already uses, so the fast path
normalizes exactly what the full encoder would. Using the newer
unicode-normalization crate changed the count for 171 code points that
gained compatibility decompositions after Unicode 9 (U+32FF, U+A7F1..).
The fast normalizer is compared with the tokenizer's for every scalar
value and on random texts.
The scanner is built without mutable state: byte_char and mapped_len replace the const table builders and the reusable mapped buffer, and iter::successors replaces the stateful piece iterator. byte_chars_match_the_byte_level_alphabet checks the byte mapping against ByteLevel for every scalar value.
Co-authored-by: yassin <yassin@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(rust_bridge): bound concurrent token-count encodes and share the Anthropic tokenizer predicate
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>
The writer health probe only ran SELECT 1, which a read-only Postgres
session answers fine, so a pooled connection left pointing at a demoted
primary kept failing every write with SQLSTATE 25006 until the pod was
restarted. Probe transaction_read_only instead, treat a 25006 on the
request path as a signal to recreate the client, and back off
exponentially while the database as a whole stays read-only so a replica
or an in-progress failover does not get its engine killed every cycle.
Co-authored-by: yassin <yassin@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Profiling the sidecar-enabled gateway at 700 rps showed ~2.4% of all samples
in get_model_group_info called per request from budget reservation, plus
get_deployment_model_info for tiered pricing tables. Both are read-only lookups
over the model list, so serve them from the Router's lru caches and clear the
deployment cache alongside the group cache when the model list changes.
The deployment-info cache is a per-router lru_cache built in __init__ rather
than a class-level decorated method, so it does not pin Router instances in a
process-wide cache and is dropped with the router.
A price data reload replaces litellm.model_cost without touching model_list, so
the reload replay also clears both caches; otherwise reservation would keep
pricing against the old catalog until an unrelated model-list change.
Co-authored-by: yassin <yassin@berri.ai>
* feat(mock): report admission-time input token count in mock_response usage
Mock completions always reported prompt_tokens=10, so spend tracking, TPM metrics, budgets and the tokens-per-minute autoscaling signal saw 10 tokens for a 100k-token request. Budget reservation now carries the admission-time input token count in the reservation record, and mock_completion reads it back so mock traffic exercises the same spend and TPM paths as real traffic without any extra tokenizer work.
* fix(mock): keep a zero admission input token count instead of falling back to 10
---------
Co-authored-by: yassin <yassin@berri.ai>
* fix(proxy): bound concurrent key and spend-counter DB lookups to stop prisma pool thrash
A cache-miss burst fanned every get_key_object DB fallback and SpendCounterReseed
point lookup into the prisma query-engine httpx pool at once. httpcore's request
assignment is O(queued x connections) per event, so the event loop spent most of its
time in pool bookkeeping and the logging worker's 20s wait_for tripped. Callers now
wait on a small shared semaphore (PROXY_DB_LOOKUP_MAX_CONCURRENCY, default 25)
instead of queueing inside httpcore
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): type the in-flight counting prisma fake
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(proxy): drop module docstring from db_lookup_gate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): type the in-flight counting table fake
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yucheng <yucheng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Encode complete non-Claude source names and include source_model in the
Claude Code listing. Preserve configured route and alias precedence,
normalize once before model policy checks, and select CLI models using
explicit source identity instead of name stripping or positional joins.
Resolves LIT-7360
Claude-Session: https://claude.ai/code/session_01WyqeRhfZGm26zAnHx9P3kq
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* fix(proxy): surface runtime-registered callbacks in /get/config/callbacks
Config-file callbacks fire at runtime but never appear in the UI Logging
and Alerts page because /get/config/callbacks only reads the DB-merged
config. Append runtime-registered callbacks from LoggingCallbackManager
as read-only rows, deduplicated against configured rows via alias
normalization. UI hides edit/delete/test actions for read-only rows.
* fix: filter internal proxy hooks from runtime callbacks, update test
- Filter _PROXY*, ShadowEval, ServiceLogging, SkillsInjection, ResponsesID prefixes
- Update test to exclude read_only rows from count assertions
- Still allows deployment/guardrail callbacks to surface if configured
Note: comprehensive internal-hook filtering deferred, live-pr-risk will
observe real behavior on running proxy.
* fix: guard non-list config callbacks in get_config, use monkeypatch in tests
- Line-concat type error: normalize_callback now returns empty list for non-list types (dict/tuple/set) instead of passing through unchanged; prevents TypeError when config values are non-list
- Test quality TQ005: replace manual try/finally save-restore of litellm.callbacks with monkeypatch.setattr in test_get_config_callbacks_appends_runtime_only_callbacks and test_get_config_callbacks_redacts_runtime_only_row_secrets_for_view_only_admin
- Ruff format: wrap _internal_callback_prefixes tuple and isinstance check across multiple lines to respect 120-char limit
- All three new tests pass
* fix: rework runtime callback inventory filtering and dedup
- Filter internal proxy hooks by name: _PROXY_ prefix plus fixed internal names (cache, _ProxyDBLogger, deployment callbacks, service hooks)
- Hide guardrail instances and runtime instances of already configured callbacks via CustomLoggerRegistry class lookup
- Sort runtime rows and dedup per mode for stable output
- normalize_callback returns tuples for str/None/list config values and empty for any other type
- Tests mock get_callbacks_by_type explicitly and pin the exact row set; UI test covers read_only action hiding
* fix: list dict-shaped callback config values by their keys
Dict-valued success_callback/failure_callback/callbacks settings previously listed their keys as editable rows; keep that behavior instead of dropping them to read-only runtime rows. Adds a pin test for the dict shape.
* fix: mark dotted-path callbacks read-only to prevent duplicate display
Configured callbacks loaded from dotted Python paths (e.g. custom_callbacks.my_logger) are never matched against runtime instances by name because the registry uses short canonical names (e.g. langsmith, arize). Mark these rows read-only to prevent the UI from attempting delete operations that would fail at the endpoint level anyway.
* fix: dedupe dotted-path callbacks by instance module instead of marking them read-only
A dotted-path callback loaded from config registers as an object, so it
surfaces at runtime under its class name and never matched the configured
string, producing a second row. Marking the config row read_only hid the
duplicate but also hid delete, which does work for these rows.
Match the live instance back to its configured entry by module and drop it
from the runtime rows, so the callback stays a single editable row.
* test: cover dotted-path dedup across success, failure, and callbacks modes
* fix(proxy): filter runtime callback inventory by object identity and label read-only rows in the UI
Runtime-only rows were filtered by callback name, which missed initialized
CustomLogger instances, router and proxy hook methods, guardrails, and
user functions. The inventory now inspects the live callback objects
through a public LoggingCallbackManager.get_callback_objects accessor
and hides litellm-internal hooks, guardrails, and instances of already
configured callbacks. The dashboard shows a Read only label for
runtime-only rows instead of an empty action cell
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): keep configured-callback assertions minimal when runtime rows are present
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): hide internal cache string callback from runtime callback inventory
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): hide auto-registered vector store hook from callback inventory
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): keep YAML OTel-family callbacks listed next to a configured one
arize, weave_otel and langfuse_otel all initialize OpenTelemetry subclasses, so hiding runtime
callbacks by configured class made one saved OTel callback swallow its YAML siblings. Match runtime
instances by their own callback_name and only fall back to class identity for bare OpenTelemetry
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): cover scalar and null YAML callback keys in callback inventory
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(proxy): drop docstrings that restate callback inventory helpers
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): keep runtime-only s3 and sqs callbacks in UI Logging inventory
_is_litellm_internal_callback checked registry membership with the display alias (s3, sqs), which is not a registry key, so runtime-only S3Logger and SQSLogger instances were classified as internal and dropped from /get/config/callbacks. Check the registered name instead and cover both loggers in the internal-exclusion regression test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Resolves the conflicts with the lite configure claude work from #40319: every persistent
writer and reader of Claude Code's settings file now resolves it through CLAUDE_CONFIG_DIR,
the lite up backup check only guards the default file, and each settings file keeps its own
undo receipt (the default file keeps ~/.litellm/claude_configure_state.json, any other file
gets ~/.litellm/claude_configure_state/<sha256 of its resolved path>.json).
Persistently route Claude Code through a LiteLLM proxy with a long-lived virtual key or the stored lite login, turn on gateway model discovery so /model lists the proxy's models, optionally pick the model Claude Code starts on, and record what changed so unconfigure restores only the keys the user has not touched since. lite login --config-claude writes through the same receipt and is undoable too. The two settings merges (lite up / --config-claude and lite autoroute) collapse into one credential-aware merge
Claude Code reads settings.json from CLAUDE_CONFIG_DIR when it is set,
while lite login --config-claude wrote to ~/.claude/settings.json and
lite claude checked that same file for its apiKeyHelper. With the
override set, lite could drop ANTHROPIC_AUTH_TOKEN because the helper
lives in a file Claude Code never reads, leaving it with no key at all.
claude_settings_path(environ) now picks the file the way Claude Code
does, and both commands go through it. The CLI test directory gets an
autouse conftest that isolates HOME, USERPROFILE, and CLAUDE_CONFIG_DIR
per test and fails any test that writes the developer's real
settings.json, restoring it first.
The FK cascade drops the LiteLLM_JWTKeyMapping row, but the cached
jwt_key_mapping:{claim}:{value} entry still resolved to the deleted token
hash, so every JWT call from that identity failed until
virtual_key_mapping_cache_ttl expired instead of auto-registering against a
recreated key. delete_verification_tokens now snapshots the mapping cache
keys before the delete and evicts them across replicas afterwards, the same
way /key/regenerate already does.
Claude-Session: https://claude.ai/code/session_011Tn3657NkV6ojLqewL64Kb
`lite login --config-claude` writes an apiKeyHelper into ~/.claude/settings.json,
and `lite claude` then also exported ANTHROPIC_AUTH_TOKEN, so Claude Code opened
with its "Both ANTHROPIC_AUTH_TOKEN and apiKeyHelper set" banner and, because the
env token wins, never ran the helper that was meant to refresh the key.
When the launch key is the stored login key and settings.json carries exactly
the helper lite wrote for this base URL, `lite claude` now leaves the token out
of the env (dropping an inherited one) so Claude Code asks the helper. An
explicit --api-key or LITELLM_PROXY_API_KEY, a helper for another proxy, a
hand-written helper, or no settings file keep the previous behavior.
Team admins can create and edit team models through /model/new, PUT /model/update
and PATCH /model/{id}/update. The proxy forwards aws_session_tags to STS under its
own identity, so a team admin could pick tags that unlock aws:PrincipalTag gated
resources. Only proxy admins may now set or change aws_session_tags there; an
unchanged tag set still passes so team admins can edit other fields
A team-scoped auto-router is stored under an internal
model_name_{team_id}_{uuid} with the caller-facing name in
model_info.team_public_model_name, and the four pre-routing strategy
registries key on that internal name. A team key asks for the public name,
so the strategy lookup missed, the team early-resolve exit handed back the
marker deployment itself, and every call 400'd with "Unmapped LLM provider".
The strategy lookup now resolves the requested name through the same
team-first, then global, then admin-across-teams deployment resolution the
deployment path uses, and looks the registries up under the model_name of
whatever that resolves to. Both exits of _common_checks_available_deployment
drop strategy markers through one helper, so a marker-only resolution is
rejected as uncallable on every path. The request team id has one reader.
Resolves LIT-7363
Claude-Session: https://claude.ai/code/session_01NU97S7d2FUDDvTk59k53Wp
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* fix(proxy): accept non-string callback vars in default_team_settings
A YAML boolean such as turn_off_message_logging: true in a
default_team_settings block failed TeamCallbackMetadata's str-only
callback_vars validation and errored the request before any callback
ran. Stringify the value the same way AddTeamCallback does.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): drop docstring from default_team_settings bool regression test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): move default_team_settings bool regression test to mapped pre_call_utils suite
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yucheng <yucheng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
