CLIProxyAPI

Add candidate count parameter to OpenAI request

extremk · 2026-01-10 18:50:13 +08:00

5bb9c2a2bd

Add candidate count handling in OpenAI request

extremk · 2026-01-10 18:49:29 +08:00

0b5bbe9234

Handle 'n' parameter for candidate count in requests

Added handling for the 'n' parameter to set candidate count in generationConfig.

extremk · 2026-01-10 18:48:33 +08:00

14c74e5e84

Add candidate count handling in OpenAI request

extremk · 2026-01-10 18:47:41 +08:00

6448d0ee7c

Enhance Gemini to OpenAI response conversion

Refactor response handling to support multiple candidates and improve parameter management.

extremk · 2026-01-10 18:46:25 +08:00

b0c17af2cf

Merge branch 'sdk' into dev

Luis Pater · 2026-01-10 16:26:23 +08:00

v6.6.98 8cfe26f10c

Merge pull request #955 from router-for-me/api

feat(codex): add subscription date fields to ID token claims

Luis Pater · 2026-01-10 16:26:07 +08:00

80db2dc254

feat(executor): add HttpRequest support across executors for better http request handling

Luis Pater · 2026-01-10 16:25:25 +08:00

e8e3bc8616

refactor(logger): remove unnecessary request details limit logic

Luis Pater · 2026-01-10 14:46:59 +08:00

v6.6.97 bc3195c8d8

feat(codex): add subscription date fields to ID token claims

hkfires · 2026-01-10 11:15:20 +08:00

6494330c6b

Fixed: #941

fix(translator): ensure fallback to valid originalRequestRawJSON in response handling

Luis Pater · 2026-01-10 01:01:09 +08:00

v6.6.96 4d7f389b69

Merge pull request #947 from pykancha/fix-memory-leak

Resolve memory leaks causing OOM in k8s deployment

Luis Pater · 2026-01-10 00:40:47 +08:00

95f87d5669

Merge pull request #938 from router-for-me/log

refactor(logging): clean up oauth logs and debugs

Luis Pater · 2026-01-10 00:02:45 +08:00

v6.6.95 c83365a349

Merge pull request #943 from ben-vargas/fix-tool-mappings

Fix Claude OAuth tool name mapping (proxy_)

Luis Pater · 2026-01-09 23:52:29 +08:00

6b3604cf2b

Fixed: #942

fix(executor): ignore non-SSE lines in OpenAI-compatible streams

Luis Pater · 2026-01-09 23:41:50 +08:00

v6.6.94 af6bdca14f

fix: Remove investigation artifacts

hemanta212 · 2026-01-09 17:47:59 +05:45

1c773c428f

Use unprefixed Claude request for translation

Keep the upstream payload prefixed for OAuth while passing the unprefixed request body into response translators. This avoids proxy_ leaking into OpenAI Responses echoed tool metadata while preserving the Claude OAuth workaround.

Ben Vargas · 2026-01-09 00:54:35 -07:00

e785bfcd12

fix(server): resolve memory leaks causing OOM in k8s deployment

- usage/logger_plugin: cap modelStats.Details at 1000 entries per model
- cache/signature_cache: add background cleanup for expired sessions (10 min)
- management/handler: add background cleanup for stale IP rate-limit entries (1 hr)
- executor/cache_helpers: add mutex protection and TTL cleanup for codexCacheMap (15 min)
- executor/codex_executor: use thread-safe cache accessors

Add reproduction tests demonstrating leak behavior before/after fixes.

Amp-Thread-ID: https://ampcode.com/threads/T-019ba0fc-1d7b-7338-8e1d-ca0520412777
Co-authored-by: Amp <amp@ampcode.com>

hemanta212 · 2026-01-09 13:33:46 +05:45

47dacce6ea

Fix Claude OAuth tool name mapping

Prefix tool names with proxy_ for Claude OAuth requests and strip the prefix from streaming and non-streaming responses to restore client-facing names.

Updates the Claude executor to:
- add prefixing for tools, tool_choice, and tool_use messages when using OAuth tokens
- strip the prefix from tool_use events in SSE and non-streaming payloads
- add focused unit tests for prefix/strip helpers

Ben Vargas · 2026-01-09 00:10:38 -07:00

dcac3407ab

build(docker): move stats export execution after image build

hkfires · 2026-01-09 11:24:00 +08:00

7004295e1d

refactor(logging): clean up oauth logs and debugs

hkfires · 2026-01-09 11:20:55 +08:00

ee62ef4745

fix(executor): handle context cancellation and deadline errors explicitly

Luis Pater · 2026-01-09 10:48:29 +08:00

ef6bafbf7e

refactor(amp): remove duplicate comments in response rewriter

Luis Pater · 2026-01-09 08:21:13 +08:00

ed28b71e87

refactor(response): enhance parameter handling for Codex to Claude conversion

Luis Pater · 2026-01-09 05:20:19 +08:00

v6.6.93 d47b7dc79a

Merge pull request #787 from sususu98/fix/antigravity-429-retry-delay-parsing

fix(antigravity): parse retry-after delay from 429 response body

Luis Pater · 2026-01-09 04:45:25 +08:00

49b9709ce5

Merge pull request #763 from mvelbaum/feature/improve-oauth-use-logging

feat(logging): disambiguate OAuth credential selection in debug logs

Luis Pater · 2026-01-09 04:43:21 +08:00

a2eba2cdf5

Merge pull request #553 from XInTheDark/fix/builtin-tools-web-search

fix(translator): preserve built-in tools (web_search) to Responses API

Luis Pater · 2026-01-09 04:40:13 +08:00

3d01b3cfe8

Merge pull request #605 from soilSpoon/feature/amp-compat

feature: Improves Amp client compatibility

Luis Pater · 2026-01-09 04:28:17 +08:00

af2efa6f7e

Merge pull request #901 from uzhao/vscode-plugin

Vscode plugin

Luis Pater · 2026-01-08 22:22:27 +08:00

d73b61d367

feat(executor): centralize systemInstruction handling for Claude and Gemini-3-Pro models

Luis Pater · 2026-01-08 21:05:33 +08:00

v6.6.92 59a448b645

Merge pull request #921 from router-for-me/atgy

fix(executor): update gemini model identifier to gemini-3-pro-preview

Chén Mù · 2026-01-08 19:20:32 +08:00

v6.6.91 4adb9eed77

fix(executor): update gemini model identifier to gemini-3-pro-preview

Update the model name check in `buildRequest` to target "gemini-3-pro-preview" instead of "gemini-3-pro" when applying specific system instruction handling.

hkfires · 2026-01-08 19:14:52 +08:00

b6a0f7a07f

feat(executor): update system instruction handling for Claude and Gemini-3-Pro models

Luis Pater · 2026-01-08 12:42:26 +08:00

v6.6.90 1b2f907671

feat(executor): add model-specific support for "gemini-3-pro" in execution and payload handling

Luis Pater · 2026-01-08 12:27:03 +08:00

bda04eed8a

feat(executor): enhance Antigravity payload with user role and dynamic system instructions

Luis Pater · 2026-01-08 10:55:25 +08:00

v6.6.89 67985d8226

Update README_CN.md

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

Jianyang Zhao · 2026-01-07 20:07:01 -05:00

cbcb061812

Update README.md

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

Jianyang Zhao · 2026-01-07 20:06:55 -05:00

9fc2e1b3c8

Add Claude Proxy VSCode to README_CN.md

Added information about Claude Proxy VSCode extension.

Jianyang Zhao · 2026-01-07 20:03:07 -05:00

3b484aea9e

Add Claude Proxy VSCode extension to README

Added Claude Proxy VSCode extension to the README.

Jianyang Zhao · 2026-01-07 20:02:50 -05:00

963a0950fa

fix(executor): remove unused tokenRefreshTimeout constant and pass zero timeout to HTTP client

Luis Pater · 2026-01-07 18:16:49 +08:00

v6.6.88 f4ba1ab910

feat(management): add PostOAuthCallback handler to token requester interface

Luis Pater · 2026-01-07 10:47:32 +08:00

v6.6.87 2662f91082

Merge pull request #888 from router-for-me/api-call-TOKEN-fix

fix(management): refresh antigravity token for api-call $TOKEN$

Luis Pater · 2026-01-07 01:19:24 +08:00

v6.6.86 c1db2c7d7c

fix(auth): error when antigravity refresh token missing during refresh

LTbinglingfeng · 2026-01-07 01:09:50 +08:00

5e5d8142f9

fix(management): refresh antigravity token for api-call $TOKEN$

LTbinglingfeng · 2026-01-07 00:14:02 +08:00

b01619b441

docs: add 9Router to community projects in README

Luis Pater · 2026-01-06 23:15:28 +08:00

v6.6.85 f861bd6a94

Merge pull request #871 from decolua/patch-1

Update README.md

Luis Pater · 2026-01-06 22:58:53 +08:00

6dbfdd140d

fix(codex): use unicode title casing for plan

zhiqing0205 · 2026-01-06 10:24:02 +08:00

aa8526edc0

feat(codex): include plan type in auth filename

zhiqing0205 · 2026-01-06 02:25:56 +08:00

ac3ca0ad8e

fix(antigravity): preserve finish_reason tool_calls across streaming chunks

When streaming responses with tool calls, the finish_reason was being
overwritten. The upstream sends functionCall in chunk 1, then
finishReason: STOP in chunk 2. The old code would set finish_reason
from every chunk, causing "tool_calls" to be overwritten by "stop".

This broke clients like Claude Code that rely on finish_reason to
detect when tool calls are complete.

Changes:
- Add SawToolCall bool to track tool calls across entire stream
- Add UpstreamFinishReason to cache the finish reason
- Only emit finish_reason on final chunk (has both finishReason + usage)
- Priority: tool_calls > max_tokens > stop

Includes 5 unit tests covering:
- Tool calls not overwritten by subsequent STOP
- Normal text gets "stop"
- MAX_TOKENS without tool calls gets "max_tokens"
- Tool calls take priority over MAX_TOKENS
- Intermediate chunks have no finish_reason

Fixes streaming tool call detection for Claude Code + Gemini models.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

MohammadErfan Jabbari · 2026-01-05 18:45:25 +01:00

fe6043aec7

Update README.md

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

decolua · 2026-01-05 20:54:33 +07:00

386ccffed4

1665 Commits