feat(proxy): forward Anthropic speed:fast to Codex priority tier#170
Merged
james-6-23 merged 1 commit intoMay 26, 2026
Conversation
📝 WalkthroughWalkthroughAdds Anthropic Messages ( ChangesAnthropic Messages Service Tier Support
Sequence DiagramsequenceDiagram
participant Client
participant TranslateAnthropicToCodex
participant upstreamServiceTier
participant CodexUpstream
participant Handler
participant UsageLogger
Client->>TranslateAnthropicToCodex: POST /v1/messages (speed)
TranslateAnthropicToCodex->>TranslateAnthropicToCodex: shouldUseCodexPriorityForAnthropicSpeed()
TranslateAnthropicToCodex->>upstreamServiceTier: upstreamServiceTier("priority")
upstreamServiceTier-->>TranslateAnthropicToCodex: priority tier value
TranslateAnthropicToCodex->>CodexUpstream: send Codex request (maybe service_tier)
CodexUpstream-->>Handler: response.completed (may include response.service_tier)
Handler->>Handler: resolveServiceTier(actualServiceTier, serviceTier)
Handler->>UsageLogger: record UsageLogInput.ServiceTier
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
caocuong2404
force-pushed
the
feat/forward-anthropic-speed-fast-to-codex-priority
branch
from
360e46e to
bd19194
Compare
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/API.md`:
- Line 312: The table cell containing the text "模型映射映射到 Codex 模型名（如 Claude →
`gpt-5.5`）" has a duplicated word "映射"; update that string to read "模型映射到 Codex
模型名（如 Claude → `gpt-5.5`）" (remove the duplicate "映射") so the description is
correct.
- Line 315: Docs state the fallback for reasoning.effort is "medium" but runtime
actually falls back to "high" when no output_config.effort is provided and
Anthropic thinking.type is not set; update the docs entry for reasoning.effort
to reflect the runtime default "high" (keeping the rest about preferring
output_config.effort, the Anthropic thinking.type/budget_tokens mapping, and
normalization to low/medium/high/xhigh intact) so documentation matches the
implementation.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 7d7f37d4-0e29-4b0e-a8d1-fa73c1bd8189
📒 Files selected for processing (5)
docs/API.mddocs/ARCHITECTURE.mdproxy/anthropic.goproxy/anthropic_test.goproxy/handler_anthropic.go
|
|
||
| | 主题 | 行为 | | ||
| |------|------| | ||
| | **模型** | 经管理台模型映射映射到 Codex 模型名（如 Claude → `gpt-5.5`） | |
There was a problem hiding this comment.
Fix duplicated wording in the model mapping description.
Line 312 contains 模型映射映射 (duplicate “映射”).
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/API.md` at line 312, The table cell containing the text "模型映射映射到 Codex
模型名（如 Claude → `gpt-5.5`）" has a duplicated word "映射"; update that string to
read "模型映射到 Codex 模型名（如 Claude → `gpt-5.5`）" (remove the duplicate "映射") so the
description is correct.
| | **模型** | 经管理台模型映射映射到 Codex 模型名（如 Claude → `gpt-5.5`） | | ||
| | **`speed`** | Anthropic 入参仅识别官方 `speed:"fast"`，并映射为 Codex `service_tier:"priority"`；其它 `speed` 值不触发 priority | | ||
| | **`service_tier`** | Anthropic 请求侧 `service_tier` 不在当前兼容范围内，不从该字段解析 fast mode | | ||
| | **`reasoning.effort`** | 优先使用 `output_config.effort`（经归一为 `low` / `medium` / `high` / `xhigh`）；否则若存在 Anthropic **`thinking.type=enabled`**，按 `budget_tokens` 粗映射档位；否则默认 **`medium`**（避免未带 `output_config` 的客户端始终吃满 `high` 推理成本） | |
There was a problem hiding this comment.
Default reasoning.effort in docs conflicts with runtime behavior.
Line 315 says the fallback is medium, but current behavior is high (see proxy/anthropic.go Line 474-479 and proxy/anthropic_test.go Line 59-61). Please align docs and implementation to the same default to avoid cost/latency surprises.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/API.md` at line 315, Docs state the fallback for reasoning.effort is
"medium" but runtime actually falls back to "high" when no output_config.effort
is provided and Anthropic thinking.type is not set; update the docs entry for
reasoning.effort to reflect the runtime default "high" (keeping the rest about
preferring output_config.effort, the Anthropic thinking.type/budget_tokens
mapping, and normalization to low/medium/high/xhigh intact) so documentation
matches the implementation.
bd19194 to
20f37ef
Compare
Map official speed:"fast" to upstream service_tier priority, omit priority when absent, and log resolved service_tier on /v1/messages for Usage admin fast stats.
caocuong2404
force-pushed
the
feat/forward-anthropic-speed-fast-to-codex-priority
branch
from
20f37ef to
0e77bee
Compare
There was a problem hiding this comment.
🧹 Nitpick comments (1)
proxy/anthropic_test.go (1)
87-127: ⚡ Quick winAdd normalization cases for
speedmapping.Line 94 currently validates only exact lowercase
"fast". Since translator logic trims and lowercases, add uppercase/whitespace variants to lock that contract in tests.✅ Proposed test additions
cases := []struct { name string field string wantTier bool }{ {"absent omits priority", "", false}, {"speed fast maps to priority", `,"speed":"fast"`, true}, + {"speed FAST maps to priority", `,"speed":"FAST"`, true}, + {"speed with spaces maps to priority", `,"speed":" fast "`, true}, {"speed standard omits priority", `,"speed":"standard"`, false}, }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@proxy/anthropic_test.go` around lines 87 - 127, The test TestTranslateAnthropicToCodex_SpeedFastMapsToCodexPriority currently only checks the exact lowercase "fast" string; add additional table-driven cases to assert that trimmed/upper/mixed-case variants also map to service_tier "priority" (e.g., fields containing `,"speed":" FAST "`, `,"speed":"Fast"`), keeping the same assertions that "speed" is not forwarded in the Codex body; update the cases slice in the test and keep the call to TranslateAnthropicToCodexWithModels and subsequent gjson.GetBytes checks unchanged.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@proxy/anthropic_test.go`:
- Around line 87-127: The test
TestTranslateAnthropicToCodex_SpeedFastMapsToCodexPriority currently only checks
the exact lowercase "fast" string; add additional table-driven cases to assert
that trimmed/upper/mixed-case variants also map to service_tier "priority"
(e.g., fields containing `,"speed":" FAST "`, `,"speed":"Fast"`), keeping the
same assertions that "speed" is not forwarded in the Codex body; update the
cases slice in the test and keep the call to TranslateAnthropicToCodexWithModels
and subsequent gjson.GetBytes checks unchanged.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 113d30b9-66b3-4a4b-b1d2-90d4e94d608e
📒 Files selected for processing (5)
docs/API.mddocs/ARCHITECTURE.mdproxy/anthropic.goproxy/anthropic_test.goproxy/handler_anthropic.go
✅ Files skipped from review due to trivial changes (2)
- docs/ARCHITECTURE.md
- docs/API.md
There was a problem hiding this comment.
🧹 Nitpick comments (1)
proxy/anthropic_test.go (1)
87-127: ⚡ Quick winAdd normalization coverage for
speedinput.Please add one case like mixed-case/whitespace (
" FAST "), since runtime mapping trims and lowercases before checkingfast.Proposed test case addition
cases := []struct { name string field string wantTier bool }{ {"absent omits priority", "", false}, {"speed fast maps to priority", `,"speed":"fast"`, true}, + {"speed fast with spaces/case maps to priority", `,"speed":" FAST "`, true}, {"speed standard omits priority", `,"speed":"standard"`, false}, }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@proxy/anthropic_test.go` around lines 87 - 127, Add a new table case in TestTranslateAnthropicToCodex_SpeedFastMapsToCodexPriority to cover normalization: include a case with field `,"speed":" FAST "` (mixed-case with surrounding whitespace) and expect wantTier true; this verifies TranslateAnthropicToCodexWithModels trims and lowercases the speed before comparing to "fast" and ensures service_tier becomes "priority" and speed is not forwarded to the Codex body.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@proxy/anthropic_test.go`:
- Around line 87-127: Add a new table case in
TestTranslateAnthropicToCodex_SpeedFastMapsToCodexPriority to cover
normalization: include a case with field `,"speed":" FAST "` (mixed-case with
surrounding whitespace) and expect wantTier true; this verifies
TranslateAnthropicToCodexWithModels trims and lowercases the speed before
comparing to "fast" and ensures service_tier becomes "priority" and speed is not
forwarded to the Codex body.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro Plus
Run ID: 120efeaa-43d9-43c3-ae6e-5c7792072f41
📒 Files selected for processing (5)
docs/API.mddocs/ARCHITECTURE.mdproxy/anthropic.goproxy/anthropic_test.goproxy/handler_anthropic.go
✅ Files skipped from review due to trivial changes (2)
- docs/ARCHITECTURE.md
- docs/API.md
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
speed:"fast"to upstream Codexservice_tier:"priority"(matches official Codex CLIServiceTier::Fastmapping).service_tierwhenspeedis absent or notfast; do not parse Anthropic request-sideservice_tier(Priority Tier is a separate capability).service_tieron/v1/messagesfor Usage admin fast-mode stats; document behavior indocs/API.mdanddocs/ARCHITECTURE.md.Test plan
go test ./proxy/... -run 'Anthropic|ServiceTier|Speed'/fastON → upstream WS body includesservice_tier: prioritydefaultEnvidence
Claude Code:
Codex upstream request log:
Admin UI Show:
Summary by CodeRabbit
New Features
Documentation