The tool_result user message built by the /v1/messages MCP loop used tuple
content, which the Messages to Chat Completions adapter silently dropped, so
non-Anthropic models re-requested the tool until the iteration cap or the
provider rejected the follow-up. Emit list content so the existing tool_result
branch translates it into a role tool message keyed by tool_call_id.
Resolves LIT-8474
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: bot_apk <apk@cognition.ai>
* feat(guardrails): speak the Straiker v3 platform API (/api/v3/detect)
The Straiker guardrail posted a webhook envelope to /api/v1/detect/webhook.
The v3 platform exposes /api/v3/detect instead, and its integration keys
(sk_agt_…) are rejected by the v1 route with an empty 401, so a tenant on
the v3 platform could not run this guardrail at all. Measured on a
customer gateway on 2026-09-17 after they rotated to a v3 key.
v3 parses the gateway's own traffic server-side, the same contract as
Straiker's unified Kong plugin. So on v3 the guardrail relays: the
request phase posts the provider body LiteLLM received (Anthropic
Messages or OpenAI chat), the response phase posts
{straiker_phase, sse, model, request}, the answer beside the request it
answers, and Straiker derives prompt, answer, agent and archetype. Both
phases also carry the flat prompt / app_response pair: a gateway-mode
integration key scores only the flat pair and an api-mode key only the
relayed body, each ignoring the other, so one payload serves whichever
key the console issued and it is one turn either way (measured on tenant
123, both key modes, 2026-09-18).
- api_version: "v1" | "v3", unset follows the key prefix, so a v3 key
needs no extra configuration. Explicit override still wins.
- The relayed body is an allowlist of provider fields. The hook sees the
client body merged with proxy state: `deployment` carries the resolved
provider credential and `proxy_server_request` the client's own
Authorization header. Neither travels. Identity survives as the
metadata subset Straiker's LiteLLM adapter reads.
- Identity never sends a proxy placeholder. `default_user_id` and the
master-key alias were being forwarded as a user and became the
session's identity on the platform.
- Headers: x-tool: litellm (ingress), x-straiker-phase, x-straiker-user,
and x-claude-code-session-id forwarded when the client sent it.
- Verdict: hookSpecificOutput.permissionDecision on the gateway envelope,
`action` on the flat one; block on block/deny, and on a non-empty
blocked_by as a backstop. A detect-mode control reads NONE.
- An error status from Straiker is now a webhook failure. LiteLLM's HTTP
client raises on any non-2xx and the retry loop caught only connection
errors, so a 401 or 503 from Straiker escaped the guardrail as an
exception and was relayed raw to the client, bypassing fail_open /
fail_closed. Retryable statuses retry; the rest are final.
- v1 is unchanged: same envelope, same X-Straiker-Webhook-Format header.
Tests: 15 new, fixtures from the request dict a hook sees on 1.98.0 and
the verdict envelopes the v3 platform returned on 2026-09-18. Each fix
was mutation-checked (handling removed, the test fails). Live: the same
eight-case battery (chat, /v1/messages, streaming, tool call; benign,
injection, PII) passes on a gateway-mode and an api-mode key, blocks at
pre_call with the tenant's block message, and lands under the declared
agent with the end user attributed.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* feat(guardrails): name the agent per application on v3 (x-s6r-agent)
One integration key can front several applications. Straiker enumerates them
as separate agents when the turn names one, which is what the unified Kong
plugin sends as x-s6r-agent. Without it every application on a gateway
collapses onto a single agent.
- Forwards a client-supplied x-s6r-agent.
- New `agent_ref` config names one agent for a route when the client sends
nothing. The client wins, matching Kong's precedence.
- Neither set: no header, and the platform derives the agent from the traffic.
Verified live on tenant 123 against an integration whose connector is
`gateway`: three distinct values minted three observed agents, and a turn
with no hint derived one from the traffic shape. An integration whose
connector is `custom-agent` declares its agent, so every turn attributes to
that one agent and the hint is ignored (agent_ref_source: attested).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* docs(guardrails): which v3 shape is scored depends on the connector, not the key mode
The earlier comment said a gateway-mode key scores only the flat pair. Re-measured
on tenant 123 across all three integration types with one injection prompt:
custom-agent connector (Add Agent) raw body ignored flat prompt scored
gateway connector raw body scored flat prompt scored
api mode raw body scored flat prompt ignored
Behaviour unchanged: the payload already carries both shapes, which is why it works
on every type. Comment only.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix(guardrails): send exactly what the unified Kong plugin sends on v3
The v3 platform parses the gateway's traffic itself and derives agent,
archetype and identity from it. The earlier commits added to the relayed
body (a flat prompt / app_response pair, source, user_name) and to the
headers (x-tool, x-straiker-phase, x-straiker-user). None of that is in
the Kong v0.12 contract, and traffic through this guardrail was not
classifying by shape the way the same traffic through Kong does. Match
Kong byte for byte and leave classification to the platform.
Request phase: the provider body, plus session_id and
original.processed.Meta.user. Response phase: {straiker_phase, sse,
model, request} plus the same two. No flat fields, no phase or user
headers, no x-tool.
Session id follows Kong's precedence: the client's x-claude-code-session-id,
then the session LiteLLM resolved, then an md5 of system prompt + first
message so a conversation that states no session still groups across its
replays.
Routing hints complete the Kong set: x-s6r-agent (client header, else
`agent_ref`), and new `client` (x-s6r-client) and `format_hint`
(x-s6r-format) config, both optional.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* test(guardrails): sort imports in the v3 session test
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): send a streamed Messages answer back in the Messages shape on v3
On a streamed /v1/messages call the proxy rebuilds the answer as a chat
completion before the post-call hook runs, and that is what the plugin put in
the response envelope's sse field. Straiker's coding-agent reader parses a
Messages answer, so a Claude Code turn relayed this way came back
coding_agent/claude with no session and zero events scored: the model's tool
calls were never screened on the response phase. Captured live on 2026-09-18
against tenant 123, a real Claude Code Bash tool call through the proxy.
The proxy's own Anthropic adapter turns the rebuilt answer back into a Messages
response when the call arrived on the anthropic_messages route, which is what a
transport relay forwards. Chat completions calls keep the chat completion shape
and a buffered Messages answer is relayed untouched.
The regression test's fixture is the chat completion the proxy actually built
for that captured turn. After the fix the same turn scores on the response
phase (session resolved, one event, the Bash tool_use block present).
* style(straiker): ruff format the v3 guardrail and its tests
* refactor(straiker): one attempt per call in the webhook retry loop
The HTTPStatusError branch added for v3 duplicated the non-200 branch and put
_post_webhook over the strict complexity ceiling. One attempt is now its own
method that returns the verdict or a failure marked retryable, and the loop only
decides whether to try again. Behaviour is unchanged: retryable statuses and
transport errors retry, everything else is final.
* fix(straiker): name Claude Code's client and agent on v3 so its session lands under one coding agent
Straiker types a gateway turn as a coding agent from the "You are Claude Code"
preamble, which only the main agent turns carry. Claude Code's title and
topic-detection sidecars have their own system prompts, so they resolved by
shape as autonomous, and because they share the session id with the main turns
the whole session was filed under Autonomous rather than under a coding agent.
Kong does not hit this because its plugin config names the client and agent on
every call.
The User-Agent (claude-cli/...) is on every call including the sidecars, so the
plugin now reads it and sends x-s6r-client: claude plus, when the route names no
agent, x-s6r-agent: "Claude (LiteLLM)". A client-supplied x-s6r-agent or the
agent_ref config still wins. Verified live on tenant 123: a real Claude Code
session now lands as one coding_agent labelled "Claude (LiteLLM)" with its turns
scored, where before it split across Autonomous.
Identity: the key's own user (email then id) now outranks the end user the
request named. LiteLLM resolves Claude Code's hashed metadata.user_id as the end
user when nothing better is set, so a per-user key was being shadowed by a
session token. The key is the authenticated principal, the way a Kong consumer
is, so it wins; the request end user is the fallback.
* refactor(straiker): build the v3 request, envelope and headers as frozen mappings
The v3 builders seeded dicts and grew them, which the type-discipline gate
counts as mutable accumulators. Each is now one expression over a tuple of
pairs, frozen with MappingProxyType, and the JSON encoder unwraps a frozen
mapping through a default. The session seed and the verdict parser no longer
rebind locals. The wire is unchanged: 36 live calls through the proxy on this
commit carry the same fields, shapes, headers and identities as before, with
no mappingproxy text in any body.
* fix(straiker): satisfy basedpyright on the v3 builders
The frozen-mapping refactor left a shadowed headers local, a Mapping handed to
an HTTP client that takes a dict, an unguarded optional response, a turn id
typed object, and a redundant isinstance on already-typed texts. No behaviour
change: 4 live calls (chat, Messages, Bedrock, injection) return 200 with the
expected verdicts on this commit.
* fix(straiker): type the v3 config fields at the initializer and keep the verbose log as JSON
The four v3 routing fields (api_version, agent_ref, client, format_hint)
travelled through the untyped kwargs passthrough, which basedpyright counts
against the budget. They are now validated through a small Pydantic model at
the initializer and passed by name.
The verbose log serialized the frozen payload with default=str, which printed
a Python repr instead of JSON once the builders returned MappingProxyType.
Every serializer now unwraps a frozen mapping first. A test asserts the logged
payload parses as JSON and carries the identity; mutating the log site back to
default=str fails it.
* fix(straiker): address review findings on the v3 relay
Text completions relay their prompt: `prompt`, `suffix`, `echo` and `best_of`
join the provider allowlist, so /v1/completions traffic is screened.
The route's `agent_ref` now outranks the caller's `x-s6r-agent` header. The
header is caller-supplied, and letting it beat a pinned route would let any key
file its traffic under another application's agent and controls. On a route
that names nothing the header still names the application, which is how
several applications enumerate behind one key.
Credentials inside `tools` and `mcp_servers` (an OpenAI `mcp` tool's `headers`,
Anthropic's `authorization_token`) are replaced with `[redacted]` before the
body leaves the proxy, on both phases and in the verbose log. Detection reads
tool names, descriptions and schemas, never these.
A 200 whose body is valid JSON but not an object now reports an invalid
schema and follows the failure policy instead of raising out of the hook.
Comments that restated a constant are gone. Tests cover each change and the
failure paths (unreadable error body, client exceptions, missing response,
unmodellable request, session seeds from Anthropic block shapes); every fix
fails its test when reverted.
* fix(straiker): scrub tool credentials one level deep, without recursion
* fix(straiker): scrub only the fields that carry a credential, never a schema
The credential set is now the three fields that actually hold one on a tools
or mcp_servers entry (headers, authorization, authorization_token), read one
level deep. A function tool whose parameter schema defines a token, headers or
api_key property is relayed exactly as sent; a test pins that, and fails
against the recursive version.
* test(straiker): use example.com identities; drop a comment that restated its branch
* fix(straiker): present a legacy completion as the chat exchange it is
Straiker scores chat on both phases of a gateway turn but has no reader for a
text_completion answer: the request phase of a /v1/completions call was
scored and the response phase was refused with 501, whether or not the call
named an agent. A completion is one user turn and one assistant turn, so both
phases now present that exchange: the prompt becomes the single user message
and the TextCompletionResponse becomes a chat completion. Measured through the
proxy on this commit, both phases return 200 and score, and the derived
session is shared between them.
The derived session seed accepts the tuple the conversion produces; the test
pins the session on both phases and fails against the list-only check. The
unreachable "parsed is None" branch is folded into the failure branch, and a
malformed tools value is shown to relay as sent.
* fix(straiker): screen a completions prompt as the text the model receives
LiteLLM's /v1/completions accepts a string, a list of strings, a list of
token ids or a list of token-id lists, and decodes token ids with the
text-davinci-003 tokenizer before calling the model. The relay now renders
the prompt the same way, one user message per prompt, so a pre-tokenized
prompt is screened as the text it stands for rather than as digit strings.
A prompt in a shape this cannot render (empty, mixed, or with no tokenizer
available) is relayed untouched instead of being replaced with something
else. Tests cover all four accepted shapes and six unrenderable ones.
* fix(straiker): seed the derived session on the preamble and the first user turn
An OpenAI chat body carries its system prompt as messages[0], and the derived
session seeded on the Anthropic `system` field plus messages[0] with no role
check. For that shape the seed was the system prompt twice and the first user
turn never counted, so every unnamed conversation behind one system prompt
collapsed into one Straiker session. The seed now takes the preamble from
wherever the API puts it (`system`, `instructions`, or a leading system or
developer message) and the first message with role `user`, else a Responses
`input` string, else `prompt`. Two conversations sharing a system prompt are
two sessions again; a replayed conversation stays one.
* fix(straiker): seed the derived session on every text block of the first turn
A user turn that opens with an image or a document block and carries its
text later seeded the session on an empty string, so two different
conversations under the same preamble shared one Straiker session. Read
every text block of the turn instead of only the first block. A plain
string or a single text block seeds exactly as before.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* test(straiker): cover the tokenizer fallback, a textless first turn and Responses instructions
Three branches of the v3 relay had no test: a token-id prompt relayed as
sent when the tokenizer cannot be fetched, a first user turn with no text
seeding the session on the preamble alone, and a Responses API body
seeding on its instructions and first input turn. Each test fails when
its branch is mutated.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): seed the derived session on the principal as well as the conversation
Straiker skips turns it has already scored for a session. The derived
session hashed the system prompt and the first user turn alone, so two
users who opened a conversation with the same words shared one session,
and the second user's copy of an attack came back as a replay: unscored
and allowed. Measured live on 2026-09-20: the first user's SSN turn was
blocked (`social_security_number`, scored=2), the second user's identical
turn was allowed (`controls: []`, replayed=2).
The principal now joins the seed. Explicit session ids, the Claude Code
header and LiteLLM's own session are unchanged.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): derive the session id with sha256 and drop comments that restated constants
The derived session now hashes the principal, and CodeQL flags MD5 over an
identity as a weak hash on sensitive data. SHA-256 truncated to the same
32 hex characters keeps the id shape. Comments that only labelled the
allowlist groups or restated a constant are removed; the two that explain
a non-obvious choice stay.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): keep a blocked conversation blocked when it is replayed
Straiker de-duplicates turns it has already scored per session and
answers a replay `allow`, whatever the first verdict was. A client that
resends a blocked request, or grows the conversation past the blocked
turn, was let through: measured on 2026-09-20, `block` then `allow,
events_replayed=2` for the same session and body, and Claude Code's
automatic retry after the 400 turned a blocked poisoned-file read into
a pass.
The guardrail now remembers, per session, a fingerprint of every
conversation it blocked (a bounded, day-long in-memory cache) and blocks
a request that repeats or extends one without asking again. A different
session with the same words is a new conversation and is scored afresh.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): scope the block memory by session or principal, never by content alone
A request with no derivable session keyed the replay memory on the
conversation fingerprint alone, so one caller's block could answer
another caller's identical request. The memory is now scoped by the
session, else by the principal, and a request with neither is not
remembered at all.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* fix(straiker): remember only a block that names a control, never one that comes from state
The replay memory kept every block, including one the platform returns
because a kill switch is engaged (`action: block` with `blocked_by: []`).
An administrator lifting the kill switch then left the conversation
refused by the remembered copy: measured on 2026-09-21, traffic stayed
blocked after `POST /inventory/agents/{id}/restore` returned `engaged:
false`.
The same words are the same attack tomorrow, so a control-named block is
still worth remembering; state is not ours to cache. The parsed verdict
now carries `blocked_by` so the two can be told apart.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Phimmasone Phonpaseuth <PhimmStraiker@users.noreply.github.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
* fix(prices): add baseten/zai-org/GLM-5.3-Fast pricing with cost tracking e2e
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): assert message instead of comment on breakdown row
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(baseten): drop the live e2e cost tracking 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>
Co-authored-by: kerry <kerry@berri.ai>
* ci: add dashboard and core smoke checks across supported Python versions
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci: tighten merge smoke harness and keep mapped test diffs additive
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci: terminate proxy on readiness timeout and use contextlib.suppress in teardown
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yuneng <yuneng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(params): carry stream_chunk_size through litellm_params instead of provider params
* test(integration): fence stream_chunk_size out of every provider request body
* test(bedrock): type parametrized stream chunk test params
* test(integration): drop the contracts manifest resurrected by the main merge
* test(bedrock): type the stream_chunk_size test helpers
* test(params): finish AGENTS.md typing pass on stream_chunk_size tests
* test(integration): drop the covers marker from the stream_chunk_size wire test
---------
Co-authored-by: shrey kharbanda <shreshth@berri.ai>
* test(autoroute): wait for a valid fuzzy selection index and cancel the prompt on driver failure
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(autoroute): read the fuzzy selection through the public InquirerPy property
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): inject the HIBP client into change_password so the breached-password test never touches the network
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(mcp): only use the 200ms read timeout in the silent mode of the transport completion test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): record HIBP requests so the ordering test asserts no lookup happened
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci(codeql): filter the weak-sensitive-data-hashing false positive on the HIBP k-anonymity lookup
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>
* feat(cost-map): add Azure Foundry pricing for gpt-6-sol and gpt-6-luna
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(cost-map): give azure/eu gpt-6-sol and gpt-6-luna full model metadata
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): drop the contracts.json manifest and the covers requirement
Groups live as a GROUPS literal in run.py, the browser expectations move next to the
browser tests, and the runner fails only on pytest failure, collection errors or a
selected file that collects zero tests. The covers marker stays registered for the
existing tests but is no longer checked. The mcp directory gets its own group
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci(integration): run mcp as its own shard with xdist and a peer proxy
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci(integration): INTEGRATION_COVERAGE=1 runs the proxy under coverage for the MCP modules
The mcp shard sets it. The proxy and its peer start under coverage run in parallel mode,
get SIGTERM after the tests so coverage flushes, and the combined text and HTML reports
land in the suite results that CircleCI already stores as artifacts
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci(integration): let the test proxy flush coverage when uvicorn re-raises SIGTERM
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): add SSE, stdio, scripted, OpenAPI and OAuth 2.1 MCP peer doubles
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): add MCP transport and access-control matrices
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): add MCP credential and OAuth flow coverage
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): add MCP LLM endpoint, accounting, guardrail, resilience and lifecycle coverage
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): stop the same-URL grant test from counting a late initialize as a leaked call and satisfy the test-tree lint
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): assert the REST denied-server listing is refused or empty
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): pin the REST denied-server listing to 403 access_denied
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: yuneng <yuneng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
PATCH /model/{id}/update rejected read-modify-write edits that resent an unchanged but dangling litellm_credential_name. Existence validation now runs only when the requested name differs from the stored one; empty string and non-admin detach rejections are unchanged
Co-authored-by: yuneng <yuneng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): revoke UI session tokens on logout and password change
Adds POST /session/logout to revoke the presented UI session key server
side (previously logout was client-side only and the key stayed valid
until expiry). Password changes now revoke the user's other UI sessions:
self-change keeps the caller's session, admin reset and onboarding claim
revoke all. The BYOK OAuth cookie auth now re-resolves the embedded key
against the DB so revoked sessions get a 401.
* fix(proxy): satisfy B008 budget and backend allowlist for /session/logout
* refactor(proxy): satisfy type-discipline budget in session_endpoints
* feat(otel): emit gen_ai.conversation.id from the caller's session id on v2 LLM spans
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep the caller's header session under missing_session_id: generate and read replayed payload session ids
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): drop only the proxy-minted session id so a caller id on the other metadata key survives
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep a replayed session id hidden when it only echoes the payload trace id
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): keep a replayed session id even when the payload trace id fell back to it
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): stop reading the replayed payload's session id, the generated marker does not survive replay
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): audit gen_ai.conversation.id on otel v2 spans through a real proxy, sink and postgres
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): keep otel conversation rigs alive for the whole session so shuffled shards do not reboot the proxy per test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): stop the audit rig proxies from probing sibling test peers for model info
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): record accepted OTLP batches in the sink instead of mutating the collector
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(otel): guard the accepted batch deque so snapshots cannot race sink appends
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: mrinal <mrinal@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: yucheng <yucheng@berri.ai>
* fix(proxy): bound auth cache invalidation publish so a wedged coordination Redis cannot stall user updates
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): resolve publish callable at call time in evict_and_broadcast
The keyword-only default bound publish_auth_cache_invalidation at
function-definition time, so tests patching the module attribute observed
zero calls. Default to None, resolve the real publisher inside the body,
and keep the keyword-shaped cache_key call the existing contract asserts
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): publish auth cache invalidations in the background so a wedged coordination Redis costs handlers nothing
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): cap in-flight auth cache invalidation publishes so a wedge cannot drain the redis pool
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>
* fix(policy_engine): keep inherited parent guardrails when a child policy condition misses
Attachment applicability now walks the policy inheritance chain, so an attached child whose own condition does not match still contributes the guardrails of its unconditional ancestors, and a non-default attachment that applies through an ancestor still suppresses default attachments. The resolver continues to skip only the chain members whose own condition fails
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(policy_engine): skip a policy's pipeline when its own condition misses
resolve_pipelines_for_context returned the pipeline of a matched policy without evaluating its own condition, so a condition-missing child admitted by the chain-aware matcher still ran its pipeline. It now mirrors resolve_policy_guardrails and drops the pipeline when the policy's own condition does not match.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(policy_engine): property test that chain matching only widens to applicable ancestors
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* feat(policy_engine): log policies admitted only through an inherited ancestor
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(policy_engine): log ancestor admissions once per attachment scan
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>
* fix(vector_stores): keep config-defined vector stores listed and read-only
Vector stores declared in config.yaml were purged from the in-memory registry by /vector_store/list because the database was treated as the only source of truth. Config-defined stores now carry is_config=True, stay in the list beside database rows, are never overwritten or evicted by database state, and reject /vector_store/new, /vector_store/update and /vector_store/delete with 400. The Admin UI renders them read-only
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show vector store source and read-only state for config-defined stores
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): audit config-owned vector stores across list, writes, search, authz, peers and redis outage
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): show a visible read-only hint in the config vector store actions menu
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: ryan <ryan@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(pricing): correct cached-token fields on realtime cost-map entries
azure/gpt-realtime-2 was the only member of the gpt-realtime-2 family priced
on one side of its cached-audio meter. Azure publishes that meter as
"gpt-realtime-2 Audio cd inp Gl 1M Tokens" at 0.4 per 1M and charges the
same rate for the write that populates the cache and the read that hits it,
so cache_creation_input_audio_token_cost lands at 4e-07, matching
azure/gpt-realtime-2.1, azure/gpt-realtime-2.1-mini and the openai
gpt-realtime-2 entry. No cost path reads that field yet, so this corrects
what get_model_info reports rather than what anything bills.
The gemini Live entries go the other way. Google's Vertex context-caching
page publishes separate supported-model lists for implicit and explicit
caching, and no Live or native-audio model is in either one. Its pricing
page prints N/A in both cached-input columns for every Gemini 2.5 Flash
Live API row, where plain 2.5 Flash and 2.5 Flash-Lite both carry real
cached prices, and the Vertex model card for the family marks context
caching not supported outright. Vertex never reports cachedContentTokenCount
on a Live session either, including for a byte-identical 7,021-token prefix
replayed across sessions minutes apart, which is well past the 2,048-token
minimum the same page sets for the Gemini 2 family.
So the 7.5e-08 on the two preview siblings priced something the provider does
not sell, and supports_prompt_caching on all three claimed a capability the
model does not have. The rate comes out. The flag is set to false rather than
removed, because get_model_info maps an absent key to None, and None is how
this map spells "nobody checked" across the 2,788 entries that omit it, where
false records the vendor's documented no. Both readers of the flag gate on
`is True`, so nothing bills or behaves differently either way.
Only the cached fields change on the two 09-2025 preview entries. Their
source field points at the Gemini API pricing page rather than the Vertex
one, so they describe a different surface with its own published limits, and
their context windows are left alone rather than assumed to match the Vertex
model card that drives the GA entry.
Tests cover all three halves: the family invariant that a cached audio read
implies an equal cached audio write, a cached count on a Live entry leaving
the bill at the fresh-input total instead of adding the old 7.5e-08, and
supports_prompt_caching answering false for all three entries while still
answering true for 2.5 Flash, so the false cannot be a swallowed lookup
error.
* fix(cost): correct gemini-live-2.5-flash-native-audio limits and capabilities
Google's model card for model ID gemini-live-2.5-flash-native-audio gives a
128K context window and 64K maximum output tokens, and marks structured
output, context caching and URL context as not supported. Its modality list
is text in and out, image in, audio in and out, and video in, with no
document input of any kind.
The entry advertised a 1M context window, an off-by-one 65535 output cap, and
three capability flags the vendor marks unsupported. Context caching is the
fourth and is handled in the cached-fields change alongside its two preview
siblings.
Both the bare id and vertex_ai/gemini-live-2.5-flash-native-audio resolve to
this single entry, so the test drives the corrected values through both.
* test(integration): cover live preview cached tokens billed at the fresh rate
Co-authored-by: Marty Sullivan <marty@martysullivan.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(cost): cite dated sources for Live entry pins and drop restating docstrings
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Marty Sullivan <marty@martysullivan.com>
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(bedrock): send json_schema as a forced tool on Claude Opus 4.7 and 4.8 Converse
Bedrock rejects outputConfig.textFormat on Opus 4.7 and 4.8 with
"output_config.format: Extra inputs are not permitted", and the AWS
model cards list structured outputs as not supported for both, so
their cost-map entries no longer claim supports_native_structured_output
and json_schema requests fall back to the json_tool_call tool.
Fixes#27846
* test(bedrock): assert Opus 4.7 and 4.8 inline the schema on Invoke, move the native case to Sonnet 4.6
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(bedrock): stream /v1/messages Invoke bytes through instead of holding them in a 1024-byte chunker
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(bedrock): apply ruff format to invoke messages stream passthrough
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(bedrock): drop drive-by reformat of existing invoke messages tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(bedrock): collect streamed chunks into a tuple in passthrough regression test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(bedrock): give the passthrough regression test a 10s first-chunk budget
* test(bedrock): type the eventstream frame helper's payload as Mapping[str, object]
* test(bedrock): take the gated byte stream's chunks as an immutable Sequence
---------
Co-authored-by: mateo <mateo@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix: repair seven regressions caught by CircleCI on main
- vertex_ai: stop treating fine-tuned endpoint ids (numeric or
vertex_ai/gemini/<id>) and gemma models as Gemini 3+, which injected
temperature=1.0 and Gemini 3 thinking config into their requests (#42465)
- cost: price Azure DALL-E 3 from its azure/<quality>/<size>/dall-e-3 rows;
it only worked through the OpenAI rows that #42435 removed
- bedrock: stream bedrock/invoke/moonshot through an OpenAI-shaped chunk
decoder; the generic decoder dropped every chunk, which the
supports_response_schema flag from #42338 un-skipped in CI
- proxy: keep the public model_group on pre-routing rejections so the Usage
page groups them under the model name, not the deployment (#41077)
- cost map: mirror the base rows' capability flags onto Bedrock regional and
cross-region copies (#42254 and later syncs)
- whitelist the new regional Bedrock rows from #42543 and #42588 for the
converse routing check, following the existing regional-row convention
* fix(model-prices): mirror capability flags onto ap-southeast-3 bedrock rows
* refactor(bedrock): tighten types on the moonshot stream decoder and its tests
* test(utils): raise the post-success hook error from a guardrail in the failure-hook regression
Since #42535 a plain logger raising inside async_post_call_success_deployment_hook
is logged and the completed request returns, so the regression added by #36657 for
"a post-success error never reaches async_post_call_failure_deployment_hook" failed
with DID NOT RAISE on every main run once #42603 revived the misc unit shard. The
raising callback is now a CustomGuardrail, the one kind of callback whose
post-success raise still propagates, which keeps the original assertions intact
* test(utils): type the guardrail's success-hook request_data as a Mapping
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* feat(logger): add shared Rust diagnostics and Python logging bridge
* feat(logger): dispatch diagnostic processing through Rust
* chore: regenerate Cargo.lock after rebase
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: allowlist bounded logging tree walkers in recursive detector
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* perf(logger): skip decoding plain access arguments
* test(logger): skip embedded-python logger test when litellm deps are absent
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style: cargo fmt
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test: expect NativeDiagnosticProcessor in the native public surface
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(stub): export NativeDiagnosticProcessor via __new__ in _native.pyi
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(tracing): rename logger crate and document host sink contract
* test(logger): cover exc, stack, and nested extras in the diagnostic filter
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(logger): keep rendered redacted line when template scan flags a key pattern
The blanket REDACTED for a changed msg/color template discarded lines
whose rendered form was already redacted by the same pipeline, e.g.
'password=%s' became 'REDACTED' instead of 'password=REDACTED'. Only
fall back to REDACTED when the rendered form did not change either,
which is where interpolation can mangle the key pattern the scrub
would otherwise see.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* ci(rust): install python deps so the logger bridge test runs
The end-to-end bridge test skipped silently when litellm's Python deps
were absent. uv sync --no-install-project installs them without a
maturin build, and PYTHONPATH makes them visible to the embedded
interpreter
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Yujong Lee <yujong@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): admit dcr_bridge envelope alongside an explicit litellm credential and mint under jwt principals
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(mcp): suppress LIT002 on concrete dict header payloads
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): mint bridge envelope for jwt mapped to a key without a user_id
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): mint and admit bridge envelopes under the master key
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(mcp): bind mapped JWT envelopes to stored key tokens
* fix(mcp): preserve master envelope scope enforcement
* fix(mcp): reject bridge minting that loses JWT restrictions
---------
Co-authored-by: joshua <joshua@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Co-authored-by: Joshua Valluru <326636767+joshua-berri@users.noreply.github.com>
* fix(utils): isolate callback errors in async_post_call_success_deployment_hook
A callback that raises inside async_post_call_success_deployment_hook no longer
fails the completed request. The exception is logged with the callback class and
call_type, the response stays as it was, and later callbacks still run. Guardrail
callbacks are exempt because raising is how a post-call guardrail blocks
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(utils): drop unrelated ruff autofixes from test_utils
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(utils): drop fastapi import from guardrail propagation regression
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(utils): cover every success deployment hook call type with a raising hook
Parametrize the unit regression over video, embedding, responses, image, rerank,
transcription, chat and anthropic messages responses and assert the failure log
names the callback and call type. Run the integration test through a real proxy
for /v1/chat/completions, /v1/embeddings, /v1/responses and /v1/videos
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(integration): move raising success hook cases into the existing callback delivery file
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>
* fix(proxy): keep the in-flight daily spend batch when shutdown cancels the flush
A daily spend batch drained from the in-memory queue was dropped for good when
the scheduler tick was cancelled by shutdown, because asyncio.CancelledError
bypasses the except Exception requeue. The flush now requeues the drained rows
on cancellation and re-raises, and each daily batch upsert runs in an
interactive transaction so a statement that already reached Postgres is rolled
back with the cancel instead of committing behind the requeue
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): requeue the cancelled daily spend batch before its rollback returns
Behind a lock the rollback of the cancelled interactive transaction only
returns once the blocked statement does, which is after the shutdown flush
has already run. The commit now runs as a shielded task so the cancelled
tick requeues the batch at once and lets the rollback finish in the
background. The final flush then finds the rows and writes them exactly once
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): give the recording db a transaction seam for the bulk upsert tests
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): route the mocked daily tag spend upsert through the transaction seam
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): restore the drained Redis tag batch when shutdown cancels its commit
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>
Eight directories the misc shard named moved to tests/unit on 2026-09-20, and one
missing path makes pytest-xdist collect [0 items] for the whole shard, which the
exit-5 tolerance turned into a green required check running nothing. The shared
Run tests step now drops a path that does not exist with a :⚠️: and runs
pytest over the rest, keeping option tokens verbatim.
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* test(e2e): pin bedrock batch create with blank S3 env vars
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(bedrock): treat blank S3 env vars as unset for batch jobs
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): trim blank S3 env gateway config
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): register blank_s3_env capability and clean gateway tempdir
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): move blank S3 env batch test to its own module
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>
The two redaction tests raised the deprecated InvalidStatusCode, which the
websockets 15 asyncio client never raises, and asserted the raw 403 close
code that the handshake refusal path replaced with 1008. The refusal path
builds its close reason from the status code alone, so there is no secret
to redact there, and the handshake refusal tests already cover the error
event and the 1008 close.
Those refusal tests only passed when run after a sibling test had imported
websockets.asyncio.client, since websockets lazy-loads its exceptions
submodule. Importing InvalidStatus, Response, and Headers from their own
submodules makes them pass in any order.
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(s3): replace colons in generated log filenames
Bedrock and Vertex AI batch file uploads use s3:// and gs:// URIs as
response ids. The shared filename sanitizer replaced slashes but kept
the scheme colon, producing log object keys that Hadoop-style consumers
reject as a relative path in an absolute URI.
Fixes#40234
* test(s3): drop docstrings flagged by review
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix: answer get_api_base for github_copilot and chatgpt without running the login flow
* refactor(get_api_base): dispatch the provider helpers with if-chains
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* fix(ollama): send PNG and JPEG images without requiring Pillow
The ollama/ completion transport imported Pillow before it looked at the image, so every image request failed with a 500 on installs without Pillow. That includes the Docker image, where Pillow is only a CI dependency
Detect PNG and JPEG from their leading bytes and pass them through untouched. Pillow is now imported only when another format has to be re-encoded as JPEG, and that case still raises the same install hint
* fix(ollama): address Greptile findings on image conversion
Catch all exceptions on Pillow import, not just ImportError, so the helpful
install hint always appears. Break a line that exceeded 120 characters
The cost map's fal_ai/fal-ai/trellis-2 entry prices its output by resolution
with output_cost_per_image_512, output_cost_per_image_1024, and
output_cost_per_image_1536, which litellm/types/utils.py types and the fal_ai
cost calculator reads, but INTENDED_SCHEMA in
test_aaamodel_prices_and_context_window_json_is_valid never allowed them, so
the test fails on main with "Additional properties are not allowed". Add the
three keys next to output_cost_per_image in the schema and in the cost-under-1
field list so a per-size image price is validated like the per-size video ones
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
The test priced gemini/gemini-3.1-flash-image-preview, which #42435 removed
from the cost map as deprecated, so the calculator had no per-token rates to
keep and the hardcoded expected value no longer matched. Price the live
gemini/gemini-3.1-flash-image row instead and derive the expected cost from
that row in litellm.model_cost, so a rate change on it cannot break the test
while a calculator that drops the map's token rates still fails it.
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* feat(ui): search users by id or email when assigning a key owner
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): label users without an email by user id in key owner picker
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): freeze merged user-filter where, format create key test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(proxy): suppress module-global patch findings in ui_view_users search test
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(proxy): keep merged user-filter where as a plain dict for prisma serialization
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): forward search param from userFilterUICall to /user/filter/ui
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(ui): mention user ID in the Create Key user picker helper text
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: jesus <jesus@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(anthropic): skip non-dict content items in beta-header and file-id helpers so malformed content lists return 400 instead of 500
Fixes#42094
Supersedes #42101
Co-authored-by: Pawan-Shahane <shahanepawan511@gmail.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): spawn the DB-less regression proxy with -P so the cwd cannot shadow the pinned checkout
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): launch the DB-less proxy via -I -c with an explicit sys.path so python 3.10 works, drop DIRECT_URL, remove restating docstrings
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(e2e): gate the self-booted DB-less proxy behind the owned_gateway opt-in the Buildkite container cannot satisfy
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(anthropic): move the bare string content item repro to tests/integration
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(tests): wrap the anthropic bare string wire test to the 120 column limit
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(tests): wrap anthropic common_utils test literals to the 120 column limit
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style: fix ruff findings in touched test files
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Pawan-Shahane <shahanepawan511@gmail.com>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(spend): return 400 from /spend/calculate for a model with no pricing row
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(spend): assert error type and param for unpriced /spend/calculate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(spend): move the repro to tests/integration
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix: alias ModelNotMappedError re-export to satisfy F401
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(utils): raise ModelNotMappedError only when the pricing row is missing
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: kerry <kerry@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): record the GenAI exception event without the removed Events API
OpenTelemetry removed opentelemetry._events in 1.44.0, so the three imports of
it broke 7 modules under litellm.integrations.otel, including the entry point.
Two things then failed quietly: with LITELLM_OTEL_V2 set the otel callback
resolved to None and nothing was exported, and with it unset the newrelic
callback was dropped as well, because that branch imports the v2 logger
ungated
Build and emit the event through the Logs API, which both lines carry. The
event name keeps riding the event.name attribute: the event_name log record
field that replaces it only exists from 1.44.0, and this package pins 1.28.0,
so the attribute is the only form both can write. It is also what the Events
API wrote, so exported events keep their shape
Emitting a plain record drops the default the Events SDK applied, so the
timestamp now falls back to time_ns() here
* style(otel): trim the event name key and regression test prose
Keep only the constraint a reader cannot infer from the code, that the
event_name record field does not exist on the pinned OpenTelemetry line
* fix(otel): export the GenAI exception event on both OpenTelemetry 1.28 and 1.44 lines
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(otel): drop the record selection comment
Co-authored-by: Pawan-Shahane <shahanepawan511@gmail.com>
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(otel): collapse the record selection conditional for ruff format
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(otel): build the record fields with a dict literal
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(otel): set the native event_name on the 1.44 log record
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(otel): spell out the record kwargs so the type gate sees each call
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(otel): suppress the version-gated kwargs for the type gate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* refactor(otel): drop the version-window prose and correct the event_name suppression reason
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* style(otel): wrap the compat test docstring to the line limit
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---------
Co-authored-by: Pawan-Shahane <shahanepawan511@gmail.com>
Co-authored-by: yucheng <yucheng@berri.ai>
Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): keep embedding cache hits aligned with request inputs
Partial hits now send only the uncached inputs to the provider and merge
fresh vectors back into their original positions. Responses whose item
count differs from the input count (one input scoring many documents)
are no longer written to the per-input cache, since a later hit would
return a single item.
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): drop mutable collection builds flagged by the type discipline gate
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): bypass embedding cache entries written before the per input cardinality check
Embedding cache entries now carry format_version and readers treat entries without it as
misses, so entries that only hold the first row of a multi row response are refetched instead
of served until their TTL expires. The provider call also receives a copy of the request kwargs
with the uncached inputs rather than mutating the caller's mapping
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(caching): assert a partial embedding cache hit becomes a full hit on repeat
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* test(caching): await pending embedding cache writes before asserting on cache hits
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): validate cached embeddings without mutating responses or request kwargs
Validate cache rows through a frozen pydantic model so import does not depend on
TypeAdapter support for ReadOnly TypedDicts, accept string embeddings, build the
merged partial hit response instead of mutating the cached one, and hand the
provider request mapping to post call hooks
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
* fix(caching): keep cache_hit and response_ms on merged partial embedding hits
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>
* test: count a zombie grandchild as gone in the migrate deploy timeout test
A SIGKILLed grandchild whose parent died in the same killpg reparents to
PID 1 or the nearest subreaper and stays a zombie until reaped, and signal 0
still succeeds on a zombie, so the timeout test read it as alive wherever
PID 1 is slow to reap or never does. The sibling test in
tests/test_litellm/proxy/db already handled that; both now share one
process_is_gone helper that reads the /proc state and reaps its own
children, with unit tests for the live, reaped, unreaped, and foreign
zombie shapes.
* test: move the pre-commit interrupt test onto the shared zombie-aware liveness helper
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
* test(proxy): make two proxy-infra tests independent of sibling-test state
Both tests read process-global state that another module in the same xdist
worker can change, so they passed or failed on shard scheduling rather than on
the behaviour they assert.
test_gateway_plus_backend_covers_full_app checked allowlist coverage against
the live route table. gateway/main.py trims routes once, inside the lifespan,
so it only ever sees what is registered at startup; a lazy feature appends its
router on demand afterwards and is never filtered. Those routes cannot be
dropped on the floor, but they do enter the assertion the moment a sibling test
warms the feature, and 110 of them sit in neither allowlist. Subtract exactly
the lazy features this process has loaded, which leaves the assertion at full
strength for every eagerly registered route.
test_real_proxy_child_auth_privacy_and_body_policy pins prisma_client to a bare
object(). litellm.max_budget is a module global that nothing restores between
tests; once a sibling leaves it above zero, user_api_key_auth takes the
global-spend branch, dereferences prisma_client.db, and the AttributeError
surfaces as HTTP 401. Pin max_budget next to the other globals the test already
controls.
* test(proxy): measure allowlist coverage in a pristine interpreter
The previous revision subtracted paths matching a loaded lazy feature's
prefixes. Those prefixes are broad enough to swallow eagerly registered routes:
31 of them, including /openai/deployments/*, /access_group/*, /cursor/* and
/mcp, which would have made a real allowlist regression invisible.
Run the coverage check in a fresh interpreter instead. No lazy feature is
loaded there, so the route table is exactly the one gateway/main.py's lifespan
trim sees, and nothing has to be subtracted for the result to be deterministic.
The probe also reports the lazy modules it loaded and its route count, so an
empty uncovered set cannot pass vacuously.
Drop _component_paths and the four allowlist constants it used; the probe
reproduces the predicate in the child process.
* feat(proxy): configurable key_alias_pattern for key generate, update, and regenerate
Adds litellm_settings.key_alias_pattern, a regex every key_alias sent to
/key/generate, /key/service-account/generate, /key/update, and
/key/{key}/regenerate has to fully match. A non-matching alias gets a 400
that names the setting and the pattern. When set, it replaces the built-in
rule enable_key_alias_format_validation turns on, and the baseline
unsafe-name check still runs first. An invalid regex fails config load.
* fix(proxy): cap key_alias length under key_alias_pattern and type the test fixtures
* style(proxy): declare key_alias_pattern with a PEP 604 union
---------
Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
