256 Commits

Author	SHA1	Message	Date
hkfires	7f1b2b3f6e	fix(thinking): improve model lookup and validation	2026-01-15 13:06:40 +08:00
hkfires	a75fb6af90	refactor(antigravity): remove hardcoded model aliases	2026-01-15 13:06:39 +08:00
hkfires	72f2125668	fix(executor): properly handle thinking application errors	2026-01-15 13:06:39 +08:00
hkfires	0b06d637e7	refactor: improve thinking logic	2026-01-15 13:06:39 +08:00
Luis Pater	94e979865e	Fixed: #897 ... refactor(executor): remove `prompt_cache_retention` from request payloads	2026-01-12 10:46:47 +08:00
hkfires	70a82d80ac	fix(codex): only override instructions in responses for OpenCode UA	2026-01-11 15:19:37 +08:00
hkfires	ac626111ac	feat(codex): add OpenCode instructions based on user agent	2026-01-11 13:36:35 +08:00
Luis Pater	e8e3bc8616	feat(executor): add HttpRequest support across executors for better http request handling	2026-01-10 16:25:25 +08:00
Luis Pater	95f87d5669	Merge pull request #947 from pykancha/fix-memory-leak ... Resolve memory leaks causing OOM in k8s deployment	2026-01-10 00:40:47 +08:00
Luis Pater	c83365a349	Merge pull request #938 from router-for-me/log ... refactor(logging): clean up oauth logs and debugs	2026-01-10 00:02:45 +08:00
Luis Pater	6b3604cf2b	Merge pull request #943 from ben-vargas/fix-tool-mappings ... Fix Claude OAuth tool name mapping (proxy_)	2026-01-09 23:52:29 +08:00
Luis Pater	af6bdca14f	Fixed: #942 ... fix(executor): ignore non-SSE lines in OpenAI-compatible streams	2026-01-09 23:41:50 +08:00
Ben Vargas	e785bfcd12	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.	2026-01-09 00:54:35 -07:00
hemanta212	47dacce6ea	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>	2026-01-09 13:33:46 +05:45
Ben Vargas	dcac3407ab	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	2026-01-09 00:10:38 -07:00
hkfires	ee62ef4745	refactor(logging): clean up oauth logs and debugs	2026-01-09 11:20:55 +08:00
Luis Pater	ef6bafbf7e	fix(executor): handle context cancellation and deadline errors explicitly	2026-01-09 10:48:29 +08:00
Luis Pater	49b9709ce5	Merge pull request #787 from sususu98/fix/antigravity-429-retry-delay-parsing ... fix(antigravity): parse retry-after delay from 429 response body	2026-01-09 04:45:25 +08:00
Luis Pater	59a448b645	feat(executor): centralize systemInstruction handling for Claude and Gemini-3-Pro models	2026-01-08 21:05:33 +08:00
hkfires	b6a0f7a07f	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.	2026-01-08 19:14:52 +08:00
Luis Pater	1b2f907671	feat(executor): update system instruction handling for Claude and Gemini-3-Pro models	2026-01-08 12:42:26 +08:00
Luis Pater	bda04eed8a	feat(executor): add model-specific support for "gemini-3-pro" in execution and payload handling	2026-01-08 12:27:03 +08:00
Luis Pater	67985d8226	feat(executor): enhance Antigravity payload with user role and dynamic system instructions	2026-01-08 10:55:25 +08:00
Luis Pater	f4ba1ab910	fix(executor): remove unused tokenRefreshTimeout constant and pass zero timeout to HTTP client	2026-01-07 18:16:49 +08:00
Luis Pater	7a77b23f2d	feat(executor): add token refresh timeout and improve context handling during refresh ... Introduced `tokenRefreshTimeout` constant for token refresh operations and enhanced context propagation for `refreshToken` by embedding roundtrip information if available. Adjusted `refreshAuth` to ensure default context initialization and handle cancellation errors appropriately.	2026-01-04 00:26:08 +08:00
Luis Pater	2a663d5cba	feat(executor): enhance payload translation with original request context ... Refactored `applyPayloadConfig` to `applyPayloadConfigWithRoot`, adding support for default rule validation against the original payload when available. Updated all executors to use `applyPayloadConfigWithRoot` and incorporate an optional original request payload for translations.	2026-01-02 00:03:26 +08:00
hkfires	3902fd7501	fix(iflow): remove thinking field from request body in thinking config handler	2026-01-01 19:40:28 +08:00
hkfires	4fc3d5e935	refactor(iflow): simplify thinking config handling for GLM and MiniMax models	2026-01-01 19:31:08 +08:00
hkfires	8bf3305b2b	fix(thinking): fallback to upstream model for thinking support when alias not in registry	2025-12-31 18:07:13 +08:00
hkfires	89db4e9481	fix(thinking): use model alias for thinking config resolution in mapped models	2025-12-31 17:09:22 +08:00
hkfires	26efbed05c	refactor(executor): remove redundant upstream model parameter from translateRequest	2025-12-30 20:20:42 +08:00
hkfires	96340bf136	refactor(executor): resolve upstream model at conductor level before execution	2025-12-30 19:31:54 +08:00
hkfires	b055e00c1a	fix(executor): use upstream model for thinking config and payload translation	2025-12-30 17:49:44 +08:00
sususu	414db44c00	fix(antigravity): parse retry-after delay from 429 response body ... When receiving HTTP 429 (Too Many Requests) responses, parse the retry delay from the response body using parseRetryDelay and populate the statusErr.retryAfter field. This allows upstream callers to respect the server's requested retry timing. Applied to all error paths in Execute, executeClaudeNonStream, ExecuteStream, CountTokens, and refreshToken functions.	2025-12-30 16:07:32 +08:00
hkfires	08ab6a7d77	feat(gemini): add per-key model alias support for Gemini provider	2025-12-30 13:27:57 +08:00
Luis Pater	50e6d845f4	feat(cliproxy): introduce global model name mappings for improved aliasing and routing	2025-12-30 08:13:06 +08:00
Luis Pater	457924828a	Merge pull request #757 from ben-vargas/fix-thinking-toolchoice-conflict ... Fix: disable thinking when tool_choice forces tool use	2025-12-28 14:04:30 +08:00
Ben Vargas	aca2ef6359	Fix: disable thinking when tool_choice forces tool use ... Anthropic API does not allow extended thinking when tool_choice is set to "any" or a specific tool. This was causing 400 errors when using features like Amp's /handoff command which forces tool_choice. Added disableThinkingIfToolChoiceForced() that removes thinking config when incompatible tool_choice is detected, applied to both streaming and non-streaming paths. Fixes router-for-me/CLIProxyAPI#630	2025-12-27 16:31:37 -07:00
Luis Pater	3a436e116a	feat(cliproxy): implement model aliasing and hashing for Codex configurations, enhance request routing logic, and normalize Codex model entries	2025-12-28 03:06:51 +08:00
leaph	6403ff4ec4	feat(iflow): add model-specific thinking configs for GLM-4.7 and MiniMax-M2.1 ... - GLM-4.7: Uses extra_body={"thinking": {"type": "enabled"}, "clear_thinking": false} - MiniMax-M2.1: Uses reasoning_split=true for OpenAI-style reasoning separation - Added preserveReasoningContentInMessages() to support re-injection of reasoning content in assistant message history for multi-turn conversations - Added ThinkingSupport to MiniMax-M2.1 model definition	2025-12-27 18:39:15 +01:00
Luis Pater	c281f4cbaf	Fixed: #747 ... fix(translators): rename and integrate `usageMetadata` as `cpaUsageMetadata` in Claude processing logic	2025-12-27 22:02:11 +08:00
Luis Pater	3ce0d76aa4	feat(usage): add import/export functionality for usage statistics and enhance deduplication logic	2025-12-26 11:49:51 +08:00
NguyenSiTrung	969c1a5b72	refactor: extract parseGeminiFamilyUsageDetail helper to reduce duplication	2025-12-24 10:22:31 +07:00
NguyenSiTrung	872339bceb	feat: add cached token parsing for Gemini API responses	2025-12-24 10:20:11 +07:00
Luis Pater	7569320770	Merge branch 'dev' into fix/antigravity-prompt-caching	2025-12-24 03:49:46 +08:00
Luis Pater	6d1e20e940	fix(claude_executor): update header logic for API key handling ... Refined header assignment to use `x-api-key` for Anthropic API requests, ensuring correct authorization behavior based on request attributes and URL validation.	2025-12-23 22:30:25 +08:00
Luis Pater	83b90e106f	refactor(antigravity): add sandbox URL constant and update base URLs routine	2025-12-23 02:47:56 +08:00
Evan Nguyen	24e8e20b59	Merge branch 'main' into fix/antigravity-prompt-caching	2025-12-21 19:43:24 +07:00
Evan Nguyen	a87f09bad2	feat(antigravity): add session ID generation and mutex for random source	2025-12-21 17:50:41 +07:00
Luis Pater	63908869f6	Merge pull request #611 from soilSpoon/feature/antigravity ... feat(antigravity): Improve Claude model compatibility	2025-12-21 16:27:29 +08:00