- Route `openai-response` source format requests to dedicated streaming and non-streaming responses handlers.
- Add upstream URL resolution for Kimi Responses API endpoints based on auth attributes.
- Normalize `openai-response` provider format to codex during thinking parameter resolution.
Closes: #5400#5571
advanced-tool-use-2025-11-20 is sent only while tool search or another
advanced tool-use feature is on the wire, or when the caller asks for it;
2.1.258 no longer attaches it to plain tool declarations (measured: 158
inline tools, no beta). A caller-supplied afk-mode-2026-01-31 is forwarded
between fast-mode and extended-cache-ttl and is an insertion boundary for
the advisor beta.
2.1.258 Haiku helpers offer the same full compression set as the main
thread and never send X-Stainless-Async. x-client-request-id is attached
only when the client's base URL is api.anthropic.com, so the helper
transport check accepts an empty value or a valid UUID and rejects only a
malformed one.
- Normalize complex constant `oneOf` and `anyOf` tool parameter schemas into equivalent enums to prevent upstream aborts.
- Escape property keys containing dots and colons during schema updates to prevent invalid path splitting.
- Identify terminal `response.incomplete` events with zero output tokens and no content as upstream failures.
Closes: #5551
- Configure upstream connection pool under antigravity.connection-pool with enabled: false by default.
- In short connection mode, set MaxIdleConnsPerHost = -1 with DisableKeepAlives = false, ensuring immediate TCP termination after response body completion without leaking Connection: close request headers.
- When pooling is explicitly enabled (enabled: true), cap idle-conn-timeout at 210s (leaving a 30s safety buffer below Google Frontend's 240s Keep-Alive cutoff) and default max-idle-conns-per-host to 2 (bounded at 100).
- Refactor TransportCache to execute CloseIdleConnections outside the mutex lock during LRU eviction and matching closes.
- Proactively evict and close idle connections on 429 quota exhaustion across Execute, ExecuteStream, and CountTokens.
- Wire hot-reload diff detection and server reload purge hooks for graceful transport pool updates.
- Ensure Gemini, AI Studio, Vertex, and Antigravity requests end with a user turn to prevent upstream errors when trailing with model content.
- Preserve trailing turns that contain a function response and skip trailing adjustments for `countTokens` actions or Claude models.
- Introduce boundary normalization helpers to validate both leading and trailing turn roles.
Closes: #5358
- Add payload detection and preparation helpers for compaction triggers and summary generation.
- Implement AES-GCM sealing and unsealing to store and restore opaque compaction capsules across turns.
- Support expanding compaction capsules into developer context for downstream turns.
- Provide summary text extraction across Gemini, Claude, and OpenAI response formats.
- Add builders for compaction non-stream JSON responses and SSE stream frames.
Closes: #5094
- Reuse coresession.ExtractSessionInfo across HTTP headers and request payloads to unify canonical session prefix namespaces with the scheduler.
- Extract hierarchical session identities in two phases: initial extraction from request headers on entry, and authoritative deep extraction once request payloads and metadata are available.
- Support Claude Code multi-level subagents (X-Claude-Code-Agent-Id, metadata.agent_id) and Codex thread fork lineages.
- Propagate SessionID and ParentSessionID across ClientRequestMetadata, UsageReporter, and coreusage.Record without root_session_id.
- Include session_id and parent_session_id in queuedUsageDetail for Home LPushUsage forwarding and Redis consumption with self-loop guards.
- Add comprehensive test coverage for canonical headers, body extraction, ghost parent elimination, and self-referential loop guards.
- Inject '# Reporting outcomes' system block for Fable 5.1 / Mythos 5.1 models matching Claude Code 2.1.258 tengu_dapper_lagoon behavior.
- Support both string-format and block-array system prompts for Fable reporting injection.
- Add reconcileClaudeCodeFableModelAfterPayload to reconcile model additions when payload rules override model or thinking.
- Reliably re-add reporting block if omitted after payload rules modify system blocks.
- Drop injected thinking.display on Fable models when payload changes thinking type from adaptive to disabled.
- Enhance payload rule path tracking with ancestor target matching in ApplyPayloadConfigWithTrackedPaths.
- Assemble x-anthropic-billing-header strictly following official 2.1.258 binary sequence: cc_version -> cc_entrypoint -> cch -> cc_workload -> cc_is_subagent -> cc_prev_req -> cc_prompt_id.
- Track upstream request-id in thread-safe bounded session continuity state and inject into subsequent turns as cc_prev_req.
- Generate and enforce RFC 4122 UUIDv4 cc_prompt_id per user prompt turn, preserving it across tool continuations while suppressing it on probes and title helpers.
- Detect subagents via agent headers and session metadata to emit cc_is_subagent=true.
- Harmonize diagnostics and continuity lifecycles to prevent uncommitted or truncated streams from corrupting state.
- Exclude x-anthropic-billing-header system blocks from sensitive-word zero-width space injection.
- Add `codex.orphan-delegation-compatibility` configuration option and mirror it to SDK configuration.
- Convert orphan Codex delegation outputs into standard user messages for requests with `X-Openai-Subagent: collab_spawn`.
- Integrate orphan delegation rewriting into OpenAI responses request handling pipeline.
Closes: #5401
- Bump default Claude Code version baseline from 2.1.220 to 2.1.258 to resolve upstream 400 version gate error
- Update Stainless SDK package version to 0.112.1 and runtime version to v26.3.0
- Refresh default device profile, billing header build hashes, and cloaking signatures
- Decouple unconfirmed client OS/Arch test assertions from host runner platform
- Update config.example.yaml and test suites across executor and helps packages
Close go-git repository handles across store operations and tests to avoid
leaked handles blocking recovery cleanup. Abort recovery before mutating the
worktree when close failures occur and retain backups when rollback fails.
- Add `ParsePluginExecutorResponseUsage` to extract token usage from non-streaming plugin responses across Claude, Gemini, Interactions, Antigravity, and OpenAI/Codex protocols.
- Add `ObservePluginExecutorStreamUsage` to observe and aggregate token usage across streaming chunks.
Closes: #5340
Track when executor calls cross an upstream transport boundary and use that
signal to keep model/provider errors from being replaced by later local
preparation, selection, or internal failures.
Mark HTTP, websocket, relay, and usage-tracked transports as upstream
attempts, while avoiding marks for local validation, logging, missing
sessions, and successful websocket handshakes before request send.
Parse relative auth expiry metadata and adjust Antigravity refresh timing.
- Treat `allowed_warning` status as allowed for shared 5h and 7d rate limit windows.
- Ensure Fable-only rejections with warning-level shared windows remain model-scoped instead of credential-scoped.
Closes: #5275
- Introduce `WebSocketResponseObserver` capability and bump plugin ABI schema version to 4.
- Forward upstream WebSocket response frames from Codex and xAI executors to configured observers.
- Wire `WebSocketResponseObserver` across API handlers and plugin host dispatchers.
Closes: #5248
Codex and Claude already emit credential-level quota watermarks on ordinary
responses. CPA used to drop them. Keep the latest watermark in memory and
return it from the management auth-file API.
Hard rule: this is observation only. It must not change scheduling, cooldown
selection, or auth-file persistence.
Snapshot, not accumulation
- QuotaState now has ObservedAt and a bounded Signals map. MarkResult fills
them from the response headers already recorded on the request.
- Signals is the current response, not a union of earlier ones. Retry-After
and "limit reached" only appear on the response that produced them; merging
across responses would keep an expired value forever.
- A response with no quota header (transport failure, 5xx, unrelated endpoint)
leaves the previous snapshot in place.
- ObservedAt is the time of the current snapshot. It advances even when the
values did not change, so a consumer can tell a fresh reading from a stale
one.
- When two model states merge, keep the newer snapshot. Do not union keys
captured at different times.
What is observed, and what is not
- One predicate, ProviderSupportsQuotaObservation, decides the provider set.
- Keep Codex and Claude. Drop Kimi, xAI/Grok, Antigravity, and the Gemini
family (gemini/vertex/aistudio): their ordinary headers are not a reliable
credential-level remaining quota.
- Count-tokens reuses the credential but is not generation traffic.
ExecuteCount sets SkipQuotaObservation so those headers cannot replace the
last generation snapshot. Cooldown and success/failure accounting still run.
Cooldown must not overwrite the last snapshot
- Observation writes only ObservedAt and Signals.
- Cooldown writes only Exceeded, Reason, NextRecoverAt, and BackoffLevel,
through applyCooldownFields. Never assign a fresh QuotaState{...} over a
live value: that would zero the snapshot on 429, Cloudflare, credential-
scope sibling updates, and cooldown clears.
- If a credential-quota cooldown is still active, MarkResult still observes
an already-present model state. It does not create scheduler state just to
record a watermark.
- .cds files persist cooldownFieldsOf(Quota) only. Restore keeps the newer
ObservedAt, so reloading cooldown cannot clobber a newer in-memory snapshot.
- cooldownQuotaEqual still ignores observation fields, so a watermark change
cannot by itself persist cooldown or move the scheduler.
- The management payload omits every cooldown field, so it cannot be mistaken
for scheduler state or wired back into scheduling.
- Manual ResetQuota still clears the full QuotaState.
Codex websocket events
- Codex WS reports quota as codex.rate_limits frames, not HTTP headers.
ParseCodexQuotaEventHeaders turns one event into the same bounded header
shape, and MergeResponseHeaders folds it into the request-scoped holder.
additional_rate_limits is accepted as an object (websocket) or an array
(/wham/usage).
- Parse only through AppendCodexAPIWebsocketResponse. The shared
AppendAPIWebsocketResponse is also used by xAI, and xAI error frames really
do carry x-ratelimit-* headers. Parsing every frame as Codex quota would
forge Codex headers into another provider's request log.
- Also capture code_review_rate_limits.
- A malformed active-limit name drops only that one header, not the window
watermarks parsed from the same event.
- The type discriminator scans a bounded frame prefix, not every byte of
every frame.
- HTTP namespaces an extra limit by short name (x-codex-bengalfox-*); WS
namespaces it by limit name (GPT-5.3-Codex-Spark). The two paths cannot
emit the same header names. The X-Codex-Additional- prefix marks the WS
origin, and snapshot replacement keeps the two spellings from piling up.
Hardening
- Reject observed values with control characters. These strings reach the
plain-text request log, and Limit-Name is upstream-controlled, so CR/LF
could forge a header line.
- When the header cap is hit, keep plan/credits/primary ahead of
additional-limit namespaces, then sort names so truncation is deterministic.
- QuotaState.Clone deep-copies Signals and is used by Auth.Clone and
ModelState.Clone.
- Token stores still serialize credential metadata only, so observation adds
no auth-file writes.
- Add bidirectional remapping for nested tool_references inside tool_search_tool_result across non-stream, SSE stream, and multi-turn message history.
- Add advisor_ and agent_toolset_ to IsClaudeServerToolType to prevent schema stripping and MCP aliasing on native Anthropic server tools.
- Wrap MCP alias restoration errors in claudeMCPAliasRestoreError with IsRequestScoped() bool to avoid cooling down healthy OAuth credentials.
- Add unit tests for tool_search_tool_result remapping, error variants, server tool recognition, and error scoping.
* fix(executor): prepend empty user turn for model-first requests targeting Gemini/Antigravity (#4959)
When forwarding sliced conversation histories or tool calls across OpenAI Responses,
OpenAI Chat Completions, Claude Messages, and native Gemini, native Gemini and Antigravity
Gemini endpoints require that conversation contents begin with a user turn.
Normalize leading turns at the executor boundary rather than the translator layer:
- Prepend an empty user turn ({"role":"user","parts":[{"text":""}]}) for Gemini, Gemini Vertex,
AI Studio, and Antigravity Gemini generation and CountTokens requests if the first turn is 'model'.
- Keep Antigravity Claude requests untouched to avoid adapter 400 errors.
- Ensure normalization runs after payload rules so payload index overrides target the original turns.
- Use no-copy GJSON inspection to keep overhead zero on valid user-first requests.
* fix(executor): inject Antigravity leading user after reasoning replay (#4959)
Replay can insert a model functionCall at contents[0] for sliced
tool-result history. Run the empty-user prepend on the final
requestPayload, after sanitize and prepareAntigravityGeminiReasoningReplayPayload.
* feat(config): add fingerprint-profile to Claude keys and auth JSON
- Add FingerprintProfile to ClaudeKey configuration struct and normalizer
- Track fingerprint-profile in config diff
- Map fingerprint-profile / fingerprint_profile to auth attributes in file and config synthesizers
- Support fingerprint-profile in Management API PatchClaudeKey and normalization
- Add Claude billing attribution string manipulation utilities in internal/util
- Document fingerprint-profile options in config.example.yaml
* feat(claude): add fingerprint policy and request-local CLI identity
- Centralize Claude fingerprint policy resolution in claude_fingerprint_policy.go
- Support stable Claude CLI identity synthesis (UUIDv5 account_uuid and SHA-256 device_id)
seeded from API keys or stable OAuth IDs, keeping access tokens isolated
- Warn on unrecognized fingerprint-profile values
* feat(claude): apply CLI fingerprint to Messages and keep API keys caller-owned
- Wire centralized fingerprint policy into Claude and Kimi executors
- Keep first-party Anthropic API keys and delegated providers caller-owned by default
- Apply Claude Code CLI wire profile (betas, metadata, diagnostics, MCP aliases)
when fingerprint-profile=claude-code-cli is configured
- Strictly align CCH signing with native Claude Code 2.1.220: only first-party
api.anthropic.com and Vertex sign dynamic CCH; third-party gateways and Kimi
receive billing header without cch= to avoid prompt cache busting
- Respect caller-owned count_tokens bodies by default while aligning CLI shape on opt-in
- Fall back to CLIProxyAPI/<version> User-Agent when caller sends no UA in caller-owned mode
- Scope custom operator header overrides accurately in caller-owned mode
- Add comprehensive test coverage for policy resolution, gateway opt-in, Kimi, and token counting
- add shared `EnsureResponsesUsageDetails` helper to patch `usage` objects with:
- `output_tokens_details.reasoning_tokens = 0`
- `input_tokens_details.cached_tokens = 0`
- for both plain JSON and SSE `data:` frames, including multi-line frames
- apply the helper to OpenAI Response format outputs in non-stream and stream paths across executors/plugins so translated payloads consistently include required usage details
- update websocket/completion payload builders to emit default `usage` detail fields for prewarm/finish responses
Closes: #4985
The word-based virtual server spans only ~2048^2 names, and plausible
real MCP server names such as file_system or web_search are valid BIP-39
word pairs. When a caller's own server matched the derived one, its tools
stopped passing through and entered alias recovery instead: they were
silently restored to an unrelated proxied tool, or failed the request
with a 500 when no semantic suffix matched.
Record untouched caller MCP tool names as identity entries in the reverse
map, skip those entries when collecting virtual servers and recovery
candidates, and forward them unchanged on an exact hit. Recording is
skipped when nothing was aliased, so an untouched request still keeps an
empty reverse map and a no-op restore path.
Also warn instead of silently forwarding an original name when the alias
space is exhausted, report an empty embedded wordlist, trace the
semantic-suffix fallback because it guesses rather than fails, and build
both alias entry points through one shared constructor so the exhaustion
tests cannot drift away from the production path.
Refs #4916
Replace high-entropy Base32 alias IDs with request-local BIP-39 English
words so weaker models are less likely to drift tool names. Keep a
two-word virtual server plus one-word tool ID, linearly probe wordlist
space on collision without self-overlap, and fall through to unambiguous
longest semantic-suffix recovery after a successful but wrong parse.
Fixes#4916
- Add `defer reporter.EnsurePublished(ctx)` to Gemini, Gemini Vertex, and AI Studio streaming goroutines so stream reporting is always finalized on exit.
- Update Gemini usage parsing to reject all-zero `usageMetadata` frames, preventing placeholder usage events from being accepted.
Closes: #4964
- Track normalized ID state (`occupied`/`preserved`) during preprocessing to avoid remapping valid existing IDs.
- Resolve collisions by appending deterministic hash-based suffixes until a free ID is found.
- Share occupancy tracking across shortening and remapping so sanitized IDs remain stable and idempotent.
Closes: #4891
Claude Code issues measured Haiku helper requests that intentionally omit the
claude-code beta and, for the minimal shape, the system field. Treat those
profiles as confirmed native clients so CPA does not cloak or rewrite them.
Accept the native metadata builder optional parent_session_id, require
platform headers only for presence and software baseline rather than exact
equality, derive timeout and version from the same header defaults the emitter
uses, and keep discriminating signals strict: exact beta allowlist, body
shape, lowercase hex CCH, and session binding. Preserve helper transport
headers, omit synthetic stream false on markerless helpers, and gate the
billing and CCH fallback on system presence so a later pipeline-attached system
prompt cannot go upstream unsigned while the measured no-system wire stays
intact.
Review and live-test follow-ups to the shared upstream transport.
Bound the cache with an LRU that closes idle connections on eviction, so
rotating a credential's proxy or supplying a per-request base transport can
no longer leak pools.
Stop deriving a pool scope from Auth.Label: it is documented as an optional
human readable label for logging and carries no uniqueness guarantee, so two
OAuth identities sharing a label would share one TCP/TLS pool. Prefer a
refresh-token digest, which stays stable across access-token rotation and is
available to refresh requests that run before any access token exists.
Replace a typed-nil *http.Transport taken from the request context. It passes
the interface nil check, so leaving it in place made http.Client fall back to
http.DefaultTransport, which advertises h2 over ALPN and breaks the
HTTP/1.1-only fingerprint.
Only widen pool limits: treat MaxIdleConns == 0 and IdleConnTimeout == 0 as
unlimited, and leave a negative MaxIdleConnsPerHost alone because that is how
an operator disables pooling.
Size the cache for large deployments. An unused entry costs under 1 KB and no
goroutines, whereas evicting a live pool forces a fresh TCP + TLS handshake,
so capacity is not the lever for bounding memory.
Execute, executeClaudeNonStream and ExecuteStream all assign the validated
original payload to the request and then translate both values. Since both
translations saw the same bytes, every Antigravity request paid for a second
full scan of the client payload. On a captured 24MB tool-history request that
second pass cost roughly 0.9s of CPU and 651MB of allocations.
Translate once when both inputs share a backing array and hand the caller an
independent duplicate, because later stages edit the working copy in place.
Payloads that genuinely differ still get two translations.
- Create a fresh uTLS/HTTP-2 connection per request instead of reusing cached connections.
- Pass request context through dialing and TLS handshake to make `RoundTrip` cancelable.
- Wrap response bodies so the underlying HTTP/2 connection is closed when the body is closed, and clean up promptly on request/response failures.
Closes: #4878
Reuse immutable request bytes during input inspection and avoid a redundant websocket clone while preserving the required outbound envelope.
Constraint: Preserve request immutability and one final WebSocket request-body allocation
Rejected: In-place JSON mutation | request buffers are shared across translation and execution stages
Confidence: high
Scope-risk: narrow
Directive: Do not retain no-copy gjson results or mutate their backing payload while results are in use
Tested: GOTOOLCHAIN=local go test ./...; targeted 8 MiB allocation benchmarks
Not-tested: Live Codex upstream network traffic
