- Fix 7 stale comments still mentioning "限制检查" in handlers/services
- Make billingModelForRestriction explicitly list channel_mapped case
- Add slog.Warn for error swallowing in ResolveChannelMapping and
needsUpstreamChannelRestrictionCheck
- Document sticky session upstream check exemption
Move the model pricing restriction check from 8 handler entry points
to the account scheduling phase (SelectAccountForModelWithExclusions /
SelectAccountWithLoadAwareness), aligning restriction with billing:
- requested: check original request model against pricing list
- channel_mapped: check channel-mapped model against pricing list
- upstream: per-account check using account-mapped model
Handler layer now only resolves channel mapping (no restriction).
Scheduling layer performs pre-check for requested/channel_mapped,
and per-account filtering for upstream billing source.
- calculateTokenCost reduced from 80 to 15 lines
- calculateCostInternal reduced from 91 to 15 lines
- Shared logic in computeTokenBreakdown + computeCacheCreationCost
- Unified rateMultiplier <= 0 protection in both paths
Replace response.BadRequest with response.ErrorFrom + infraerrors.BadRequest
to provide machine-readable reason codes (VALIDATION_ERROR, INVALID_CHANNEL_ID,
MISSING_PARAMETER) for frontend i18n support.
- Extract newEmptyChannelCache() factory to deduplicate map init
- Extract expandPricingToCache() for model pricing expansion
- Extract expandMappingToCache() for model mapping expansion
- buildCache reduced from 110 to 50 lines
- PricingSourceChannel/LiteLLM/Fallback for resolver source
- MediaTypeImage/Video/Prompt for result.MediaType
- Reuse BillingModeToken/BillingModeImage for billing mode
- Reuse BillingModelSourceChannelMapped/PlatformAnthropic in handler
Instead of hardcoding BillingMode="image" when ImageCount>0,
let cost.BillingMode (set by CalculateCostUnified/CalculateImageCost)
take priority. This ensures channel token pricing shows "token" mode.
When ImageCount > 0, check if channel has token pricing configured:
- YES (source=channel, mode=token) → use token billing with image_output_tokens
- NO → fall back to CalculateImageCost (original per-image billing)
This allows channels to configure $/MTok pricing for image generation
models while maintaining backward compatibility for setups without
channel pricing.
- Backend: reject intervals with all-null price fields on save
- Backend: filterValidIntervals skips empty intervals in pricing resolver
- Frontend: red border + asterisk on empty interval rows
- Backend: antigravity groups now match anthropic/gemini channel pricing
Antigravity platform serves both Claude and Gemini models, but channel
pricing/mapping is configured under Anthropic/Gemini tabs. The cache
builder was using strict platform equality, causing antigravity groups
to miss all channel pricing entries, resulting in $0 billing.
Add isPlatformPricingMatch() to treat antigravity as superset of
anthropic+gemini for pricing and mapping cache indexing.
- Show channel + account mapping steps using model_mapping_chain field
- Add model_mapping_chain to AdminUsageLog TypeScript type
- Fallback to two-level display when chain is not available
- Fix cost nil guard in Anthropic/Antigravity RecordUsage paths
- Bump version to 0.1.105.31
- Fix errcheck: defer rows.Close() with nolint
- Fix errcheck: type assertion with ok check in channel cache
- Fix staticcheck ST1005: lowercase error string
- Fix staticcheck SA5011: nil check cost before use in openai gateway
- Fix gofmt: format chatcompletions_to_responses.go
- Parse candidatesTokensDetails from Gemini API to separate image/text output tokens
- Add image_output_tokens and image_output_cost to usage_log (migration 089)
- Support per-image-token pricing via output_cost_per_image_token from model pricing data
- Channel pricing ImageOutputPrice override works in token billing mode
- Auto-fill image_output_price in channel pricing form from model defaults
- Add "channel_mapped" billing model source as new default (migration 088)
- Bills by model name after channel mapping, before account mapping
- Fix channel cache error TTL sign error (115s → 5s)
- Fix Update channel only invalidating new groups, not removed groups
- Fix frontend model_mapping clearing sending undefined instead of {}
- Credits balance precheck via shared AccountUsageService cache before injection
- Skip credits injection for accounts with insufficient balance
- Don't mark credits exhausted for "exhausted your capacity on this model" 429s
Allow redeem codes with negative values to enable refund scenarios:
- Balance: negative value deducts balance (clamped to 0, never negative)
- Concurrency: negative value reduces concurrency (clamped to 0)
- Subscription: negative validity_days reduces remaining days; if
remaining days <= 0, the subscription is canceled (set to expired)
All deductions generate standard redeem code records for audit trail.
Cover isAntigravityInternalServerError body matching,
applyInternal500Penalty tier escalation, handleInternal500RetryExhausted
nil-safety and error handling, and resetInternal500Counter paths.
Move constants, detection, and penalty functions from
antigravity_gateway_service.go to antigravity_internal500_penalty.go.
Fix gofmt alignment and replace hardcoded duration strings with
constant references.
shouldMarkCreditsExhausted was blocked by isURLLevelRateLimit check when
credit overages retry returned "Resource has been exhausted (e.g. check quota).",
causing credits to never be marked as exhausted. This led to an infinite loop
where each request injected credits, bypassed model rate limits, and failed again.
- Remove isURLLevelRateLimit guard from shouldMarkCreditsExhausted (only called
for credit retry responses — if credits retry fails, mark exhausted)
- Add "resource has been exhausted" to creditsExhaustedKeywords
- Update tests to match corrected behavior
## Problem
When a proxy is unreachable, token refresh retries up to 4 times with
30s timeout each, causing requests to hang for ~2 minutes before
failing with a generic 502 error. The failed account is not marked,
so subsequent requests keep hitting it.
## Changes
### Proxy connection fast-fail
- Set TCP dial timeout to 5s and TLS handshake timeout to 5s on
antigravity client, so proxy connectivity issues fail within 5s
instead of 30s
- Reduce overall HTTP client timeout from 30s to 10s
- Export `IsConnectionError` for service-layer use
- Detect proxy connection errors in `RefreshToken` and return
immediately with "proxy unavailable" error (no retries)
### Token refresh temp-unschedulable
- Add 8s context timeout for token refresh on request path
- Mark account as temp-unschedulable for 10min when refresh fails
(both background `TokenRefreshService` and request-path
`GetAccessToken`)
- Sync temp-unschedulable state to Redis cache for immediate
scheduler effect
- Inject `TempUnschedCache` into `AntigravityTokenProvider`
### Account failover
- Return `UpstreamFailoverError` on `GetAccessToken` failure in
`Forward`/`ForwardGemini` to trigger handler-level account switch
instead of returning 502 directly
### Proxy probe alignment
- Apply same 5s dial/TLS timeout to shared `httpclient` pool
- Reduce proxy probe timeout from 30s to 10s
Replace direct row rendering with @tanstack/vue-virtual. The table
now only renders visible rows (~20) via padding <tr> placeholders,
eliminating the rendering bottleneck when displaying 100+ rows with
heavy cell components.
Key changes:
- DataTable.vue: integrate useVirtualizer (always-on), virtual row
template with measureElement for variable row heights, defineExpose
virtualizer/sortedData for external access, overflow-y/flex CSS
- useSwipeSelect.ts: dual-mode support via optional
SwipeSelectVirtualContext — data-driven row index lookup and
selection range when virtualizer is present, original DOM-based
path preserved for callers that don't pass virtualContext
Update model mapping target for claude-haiku-4-5 and
claude-haiku-4-5-20251001 from claude-sonnet-4-5 to claude-sonnet-4-6.
Includes migration script, default constants, and test updates.
TestConnection now reuses antigravityRetryLoop instead of a standalone
HTTP loop, gaining credits overages, smart retry, and 429/503 backoff
for free. AccountSwitchError is caught and surfaced as a friendly
message. Also populates RateLimitedModel in TempUnscheduled switch error.
Test fixes:
- Use RATE_LIMIT_EXCEEDED in 503 short-delay test to avoid 60x1s timeout
- Clamp waitDuration=0 instead of 999s to avoid 15s max-wait timeout
- Enhance mockSmartRetryUpstream with repeatLast and body caching
The expanded user breakdown rows already align with the parent table
columns (Requests, Token, Actual, Standard), so the repeated sub-header
wastes vertical space. Remove the <thead> from UserBreakdownSubTable.
Click on a group name, model name, or endpoint name in the distribution
tables to expand and show per-user usage breakdown (requests, tokens,
actual cost, standard cost).
Backend: new GET /admin/dashboard/user-breakdown API with group_id,
model, endpoint, endpoint_type filters.
Frontend: clickable rows with expand/collapse sub-table in all three
distribution charts.
The usage stats query defaults to a 30-day rolling window, but the
UI label said "Total"/"累计" implying lifetime aggregation. Rename
to "Last 30d"/"近30天" so the label matches the actual query semantics.
Closes#1060
The updateRateLimitUsageScript Lua script previously performed
unconditional HINCRBYFLOAT on all usage counters without checking
whether the rate limit window had expired. This caused usage to
accumulate across window boundaries in Redis while the DB correctly
reset on expiration, leading to incorrect 429 rate limiting that
could persist for up to 24 hours.
The Lua script now checks each window timestamp before incrementing:
- If the window has expired, usage is reset to the current cost and
the window timestamp is updated (matching DB-side semantics)
- If the window is still valid, usage is accumulated normally
This also resolves the async race condition where stale HINCRBYFLOAT
tasks from the worker queue could pollute a freshly rebuilt cache
after invalidation, since the script now self-corrects expired windows.
Closes#1049
- Fix gofmt alignment in admin_service.go and trailing newline in
antigravity_credits_overages.go
- Suppress errcheck for fmt.Sscanf in client.go GetMinimumAmount
Remove SimulateClaudeMaxEnabled field and related logic from
admin_service.go, and remove applyClaudeMaxCacheBillingPolicyToUsage,
applyClaudeMaxNonStreamingRewrite, setupClaudeMaxStreamingHook calls
from antigravity_gateway_service.go. These symbols are not yet
available in upstream/main.
Replace process-memory sync.Map + per-model runtime state with a single
"AICredits" key in model_rate_limits, making credits exhaustion fully
isomorphic with model-level rate limiting.
Scheduler: rate-limited accounts with overages enabled + credits available
are now scheduled instead of excluded.
Forwarding: when model is rate-limited + credits available, inject credits
proactively without waiting for a 429 round trip.
Storage: credits exhaustion stored as model_rate_limits["AICredits"] with
5h duration, reusing SetModelRateLimit/isRateLimitActiveForKey.
Frontend: show credits_active (yellow ⚡) when model rate-limited but
credits available, credits_exhausted (red) when AICredits key active.
Tests: add unit tests for shouldMarkCreditsExhausted, injectEnabledCreditTypes,
clearCreditsExhausted, and update existing overages tests.
Introduce OAuthRefreshAPI as the single entry point for all OAuth token
refresh operations, eliminating the race condition where background
refresh and inline refresh could simultaneously use the same
refresh_token (fixes#1035).
Key changes:
- Add OAuthRefreshExecutor interface extending TokenRefresher with CacheKey
- Add OAuthRefreshAPI.RefreshIfNeeded with lock → DB re-read → double-check flow
- Add ProviderRefreshPolicy / BackgroundRefreshPolicy strategy types
- Simplify all 4 TokenProviders to delegate to OAuthRefreshAPI
- Rewrite TokenRefreshService.refreshWithRetry to use unified API path
- Add MergeCredentials and BuildClaudeAccountCredentials helpers
- Add 40 unit tests covering all new and modified code paths
The upstream Gemini API returns "Insufficient account balance" which
doesn't contain the substring "insufficient balance". Add explicit
match for the full phrase to ensure the filter works correctly.
- Add 5th error filter switch IgnoreInsufficientBalanceErrors to suppress
upstream insufficient balance / insufficient_quota errors from ops log
- Extract hardcoded error strings into package-level constants for
shouldSkipOpsErrorLog, normalizeOpsErrorType, classifyOpsPhase, and
classifyOpsIsBusinessLimited
- Define ErrNoAvailableAccounts sentinel error and replace all
errors.New("no available accounts") call sites
- Update tests to use require.ErrorIs with the sentinel error
Previously, v-model.number produced "" when input was cleared, causing
JSON decode errors on the backend. Also, normalizeLimit treated 0 as
"unlimited" which prevented setting a zero quota. Now "" is converted
to null (unlimited) in frontend, and 0 is preserved as a valid limit.
ClosesWei-Shaw/sub2api#1021
When Redis cache is populated from DB with a NULL window_1d_start, the
Lua increment script only updates usage counters without setting window
timestamps. IsWindowExpired(nil) previously returned false, so the
accumulated usage was never reset across time windows, effectively
turning usage_1d into a lifetime counter. Once this exceeded
rate_limit_1d the key was incorrectly blocked with "日限额已用完".
FixesWei-Shaw/sub2api#1022
Increase MAX(bucket_start) query timeout from 3s to 5s to reduce
timeout-induced fallbacks. Shrink backfill window from 30 days to
1 hour so that fallback recomputation stays lightweight instead of
scanning the entire retention range.
The 403 detection PR changed the 401 handler condition from
`account.Type == AccountTypeOAuth` to
`account.Type == AccountTypeOAuth && account.Platform == PlatformOpenAI`,
which accidentally excluded Gemini OAuth from the temp-unschedulable path.
Fix: use `!= PlatformAntigravity` instead, preserving Gemini behavior
while correctly excluding Antigravity (whose 401 is handled by
applyErrorPolicy's temp_unschedulable_rules).
Update tests to reflect Antigravity's new 401 semantics:
- HandleUpstreamError: Antigravity OAuth 401 now uses SetError
- CheckErrorPolicy: Antigravity 401 second hit stays TempUnscheduled
- DB fallback: split into Gemini (escalates) and Antigravity (stays temp)
Add group_id and validity_days fields to CreateAndRedeemCodeRequest,
enabling subscription-type redemption codes to be created and redeemed
in a single API call.
- Type defaults to "balance" when omitted for backward compatibility
- Subscription type requires group_id (non-nil) and validity_days (>0)
- Existing balance/concurrency callers are unaffected
Backend:
- Detect and classify 403 responses into three types:
validation (account needs Google verification),
violation (terms of service / banned),
forbidden (generic 403)
- Extract verification/appeal URLs from 403 response body
(structured JSON parsing with regex fallback)
- Add needs_verify, is_banned, needs_reauth, error_code fields
to UsageInfo (omitempty for zero impact on other platforms)
- Handle 403 in request path: classify and permanently set account error
- Save validation_url in error_message for degraded path recovery
- Enrich usage with account error on both success and degraded paths
- Add singleflight dedup for usage requests with independent context
- Differentiate cache TTL: success/403 → 3min, errors → 1min
- Return degraded UsageInfo instead of HTTP 500 on quota fetch errors
Frontend:
- Display forbidden status badges with color coding (red for banned,
amber for needs verification, gray for generic)
- Show clickable verification/appeal URL links
- Display needs_reauth and degraded error states in usage cell
- Add Antigravity tier label badge next to platform type
Tests:
- Comprehensive unit tests for classifyForbiddenType (7 cases)
- Unit tests for extractValidationURL (8 cases including unicode escapes)
- Integration test for FetchQuota forbidden path
Add a dedicated modal in group management for viewing, adding, editing,
and deleting per-user rate multipliers within a group.
Backend:
- GET /admin/groups/:id/rate-multipliers - list entries with user details
- PUT /admin/groups/:id/rate-multipliers - batch sync (full replace)
- DELETE /admin/groups/:id/rate-multipliers - clear all entries
- Repository: GetByGroupID, SyncGroupRateMultipliers methods on
user_group_rate_multipliers table (same table as user-side rates)
Frontend:
- New GroupRateMultipliersModal component with:
- User search and add with email autocomplete
- Editable rate column with local edit mode (cancel/save)
- Batch adjust: multiply all rates by a factor
- Clear all (local operation, requires save to persist)
- Pagination (10/20/50 per page)
- Platform icon with brand colors in group info bar
- Unsaved changes indicator with revert option
- Unit tests for all three backend endpoints
Two SSE scanners in openai_gateway_messages.go were hardcoded to 1MB
while all other scanners use defaultMaxLineSize (500MB) with config
override. This caused Responses API streams to fail on large payloads.
QuotaBadge components (daily/weekly/total) were wrapped in a
horizontal flex container, making them visually inconsistent with
other capacity badges (concurrency, window cost, sessions, RPM)
which stack vertically. Remove the wrapper so all badges align
consistently.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When the service restarts, concurrency slots from the old process
remain in Redis, causing phantom occupancy. On startup, scan all
concurrency sorted sets and remove members with non-current process
prefix, then clear orphaned wait queue counters.
Uses Go-side SCAN to discover keys (compatible with Redis client
prefix hooks in tests), then passes them to a Lua script for
atomic member-level cleanup.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Embedded pages (purchase subscription, custom pages) now receive the
current user locale through a `lang` URL parameter, allowing iframe
content to match the user's language preference.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
erio