Show endpoint, model mapping, and request type in the ops error log
table and detail modal:
- Endpoint column with inbound/upstream tooltip
- Model column showing requested→upstream mapping with arrow
- Request type badge (sync/stream/ws) in status column
- New detail cards for inbound endpoint, upstream endpoint, request type
Mirror the existing /v1/messages platform split pattern:
- OpenAI groups → OpenAIGateway handlers (existing, unchanged)
- Non-OpenAI groups → Gateway handlers (new Anthropic-upstream path)
Updated both /v1 prefixed routes and non-prefixed alias routes
(/responses, /chat/completions). WebSocket route (/v1/responses GET)
remains OpenAI-only as Anthropic has no WebSocket equivalent.
New HTTP handlers for Anthropic platform groups accepting OpenAI-format
endpoints:
- GatewayHandler.Responses: /v1/responses for non-OpenAI groups
- GatewayHandler.ChatCompletions: /v1/chat/completions for non-OpenAI groups
Both handlers include:
- Claude Code only restriction (403 reject when claude_code_only enabled,
since these endpoints are never Claude Code clients)
- Full auth → billing → user/account concurrency → failover loop
- Ops error/endpoint context propagation
- Async usage recording via worker pool
Error responses use each endpoint's native format (Responses API format
for /v1/responses, CC format for /v1/chat/completions).
New forwarding methods on GatewayService for Anthropic platform groups:
- ForwardAsResponses: accept Responses body → convert to Anthropic →
forward to upstream → convert response back to Responses format.
Supports both streaming (SSE event-by-event conversion) and buffered
(accumulate then convert) response modes.
- ForwardAsChatCompletions: chain CC→Responses→Anthropic for request,
Anthropic→Responses→CC for response. Streaming uses dual state machine
chain with [DONE] marker.
Both methods reuse existing GatewayService infrastructure:
buildUpstreamRequest, Claude Code mimicry, cache control enforcement,
model mapping, and return UpstreamFailoverError for handler-level retry.
Show upstream model mapping (requested -> upstream) in UsageTable with arrow notation. Add requested/upstream/mapping source toggle to ModelDistributionChart with lazy loading — only fetches data when user switches tab, with per-source cache invalidation on filter changes. Include upstream_model column in Excel export and i18n for en/zh.
Cover IsValidModelSource/NormalizeModelSource, resolveModelDimensionExpression SQL expressions, invalid model_source 400 responses on both GetModelStats and GetUserBreakdown, upstream_model in scan/insert SQL mock expectations, and updated passthrough/billing test signatures.
Add model_source query parameter to GetModelStats and GetUserBreakdown handlers with explicit IsValidModelSource validation. Include model_source in cache key to prevent cross-source cache hits. Expose upstream_model in usage log DTO with omitempty semantics.
Support querying model statistics by 'requested', 'upstream', or 'mapping' dimension. Add resolveModelDimensionExpression for safe SQL expression generation, IsValidModelSource whitelist validator, and NormalizeModelSource fallback. Repository persists and scans upstream_model in all insert/select paths.
Propagate UpstreamModel through ForwardResult and OpenAIForwardResult in Anthropic direct, API-key passthrough, Bedrock, and OpenAI gateway flows. Extract optionalNonEqualStringPtr and optionalTrimmedStringPtr into usage_log_helpers.go. Store upstream_model only when it differs from the requested model.
Also introduces anthropicPassthroughForwardInput struct to reduce parameter count.
Add nullable VARCHAR(100) column to record the actual model sent to upstream providers when model mapping is applied. NULL means no mapping — the requested model was used as-is.
Includes migration, concurrent index for aggregation queries, Ent schema regeneration, and migration README correction (forward-only runner, not goose).
- Simplify OpenAI rendering: always fetch /usage, prefer fetched data over
codex snapshot (snapshot serves as loading placeholder only)
- Remove dead code: preferFetchedOpenAIUsage, isOpenAICodexSnapshotStale,
and unreachable template branch
- Add today-stats support for key accounts (req/tokens/A/U badges)
- Use formatCompactNumber for consistent number formatting
- Add A/U badge titles for clarity
- Filter zero-value window stats in UsageProgressBar to avoid empty badges
- Update tests to match new fetched-data-first behavior
- Pass todayStats/todayStatsLoading to AccountUsageCell for key accounts
- Propagate usageManualRefreshToken to force usage reload on explicit refresh
- Refresh today stats when toggling usage/today_stats columns visible
Removes hasMeaningfulWindowStats guard so the /usage endpoint consistently
returns WindowStats for both time windows. The frontend now controls
zero-value display filtering at the component level.
- Apply InboundEndpointMiddleware to all gateway route groups
- Replace normalizedOpenAIInboundEndpoint/normalizedOpenAIUpstreamEndpoint and normalizedGatewayInboundEndpoint/normalizedGatewayUpstreamEndpoint with GetInboundEndpoint/GetUpstreamEndpoint
- Remove 4 old constants and 4 old normalization functions (-70 lines)
- Migrate existing endpoint normalization test to new API
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Introduce endpoint.go with shared constants, NormalizeInboundEndpoint, DeriveUpstreamEndpoint, InboundEndpointMiddleware, and context helpers. This replaces the two separate normalization implementations (OpenAI and Gateway) with a single source of truth. Includes comprehensive test coverage.
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Extend RecordUsageInput and RecordUsageLongContextInput structs with InboundEndpoint and UpstreamEndpoint so that Claude, Gemini, and Sora handlers can record endpoint info alongside OpenAI handlers.
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Replace t.Add(24*time.Hour - time.Nanosecond) with t.AddDate(0, 0, 1) and use SQL < instead of <= for end-of-day boundaries. This avoids edge-case misses around DST transitions.
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
- apply default mapped model only when scheduling fallback is actually used
- preserve reasoning in OpenAI-compatible output via reasoning_content and avoid invalid input function_call ids
