- Added trimming for stream input lines to prevent processing of unnecessary whitespace.
- Improved handling of unsupported prefixes and malformed JSON responses, ensuring errors are recorded and propagated appropriately.
Fixed: #2690
- Moved `ApplyThinking` logic earlier in `openai_compat_executor` to align with configuration application sequence.
- Added test to verify payload override precedence over Thinking suffix configuration.
All streaming executors use bare channel sends (out <- chunk) inside goroutines
that process upstream SSE responses. When the downstream consumer disconnects
(client timeout, network drop, etc.), these sends block indefinitely, causing
the goroutine and all associated resources (HTTP response body, scanner buffers,
translation state) to leak permanently.
Over time, leaked goroutines accumulate monotonically, leading to RSS growth
from ~30MB to 3.7GB+ and eventual OOM kills on resource-constrained VPS hosts.
Fix: Replace all bare 'out <- ...' sends with:
select {
case out <- ...:
case <-ctx.Done():
return
}
This ensures goroutines terminate promptly when the request context is canceled,
allowing GC to reclaim all associated resources.
Affected executors (9 files, 36+ send sites):
- antigravity_executor.go (5 sites)
- gemini_cli_executor.go (6 sites)
- gemini_vertex_executor.go (6 sites)
- aistudio_executor.go (4 sites)
- gemini_executor.go (3 sites)
- openai_compat_executor.go (3 sites)
- claude_executor.go (4 sites)
- codex_executor.go (2 sites)
- kimi_executor.go (3 sites)
- Introduced `DisableImageGenerationMode` with support for `false`, `true`, and `chat` values.
- Updated payload handling to preserve `image_generation` on images endpoints when `chat` mode is enabled.
- Modified OpenAI image handlers (`ImagesGenerations`, `ImagesEdits`) to respect tri-state logic.
- Added unit tests for `DisableImageGenerationMode` behavior and endpoint-specific handling.
- Enhanced configuration diff logging to support `DisableImageGenerationMode`.
- Introduced a `Disabled` flag to OpenAI compatibility configurations.
- Updated routing, auth selection, and API handling logic to respect the `Disabled` state.
- Extended relevant APIs, YAML configurations, and data structures to include the `Disabled` field.
- Adjusted all relevant loops and filters to skip disabled providers.
Closes: #3060#3059#2977
- Replaced all instances of `bytes.Clone` with direct references to enhance efficiency.
- Simplified payload handling across executors and translators by eliminating unnecessary data duplication.
Update ApplyThinking signature to accept fromFormat and toFormat parameters
instead of a single provider string. This enables:
- Proper level-to-budget conversion when source is level-based (openai/codex)
and target is budget-based (gemini/claude)
- Strict budget range validation when source and target formats match
- Level clamping to nearest supported level for cross-format requests
- Format alias resolution in SDK translator registry for codex/openai-response
Also adds ErrBudgetOutOfRange error code and improves iflow config extraction
to fall back to openai format when iflow-specific config is not present.
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.
Expose thinking/effort normalization helpers from the executor package
so conversion tests use production code and stay aligned with runtime
validation behavior.
When using OpenAI-compatible providers with model aliases (e.g., glm-4.6-zai -> glm-4.6),
the alias resolution was correctly applied but then immediately overwritten by
ResolveOriginalModel, causing 'Unknown Model' errors from upstream APIs.
This fix skips the ResolveOriginalModel override when a model alias has already
been resolved, ensuring the correct model name is sent to the upstream provider.
Co-authored-by: Amp <amp@ampcode.com>
- Added support for parsing and normalizing dynamic thinking model suffixes.
- Centralized budget resolution across executors and payload helpers.
- Retired legacy Gemini-specific thinking handlers in favor of unified logic.
- Updated executors to use metadata-based thinking configuration.
- Added `ResolveOriginalModel` utility for resolving normalized upstream models using request metadata.
- Updated executors (Gemini, Codex, iFlow, OpenAI, Qwen) to incorporate upstream model resolution and substitute model values in payloads and request URLs.
- Ensured fallbacks handle cases with missing or malformed metadata to derive models robustly.
- Refactored upstream model resolution to dynamically incorporate metadata for selecting and normalizing models.
- Improved handling of thinking configurations and model overrides in executors.
- Removed hardcoded thinking model entries and migrated logic to metadata-based resolution.
- Updated payload mutations to always include the resolved model.
Introduce `PayloadConfig` in the configuration to define default and override rules for modifying payload parameters. Implement `applyPayloadConfig` and `applyPayloadConfigWithRoot` to apply these rules across various executors, ensuring consistent parameter handling for different models and protocols. Update all relevant executors to utilize this functionality.
Add `ensurePublished` to guarantee request counting even when usage fields (e.g., tokens) are absent in OpenAI-compatible executor responses, particularly for streaming paths.
feat(translator): add token counting functionality for Gemini, Claude, and CLI
- Introduced `TokenCount` handling across various Codex translators (Gemini, Claude, CLI) with respective implementations.
- Added utility methods for token counting and formatting responses.
- Integrated `tiktoken-go/tokenizer` library for tokenization.
- Updated CodexExecutor with token counting logic to support multiple models including GPT-5 variants.
- Refined go.mod and go.sum to include new dependencies.
feat(runtime): add token counting functionality across executors
- Implemented token counting in OpenAICompatExecutor, QwenExecutor, and IFlowExecutor.
- Added utilities for token counting and response formatting using `tiktoken-go/tokenizer`.
- Integrated token counting into translators for Gemini, Claude, and Gemini CLI.
- Enhanced multiple model support, including GPT-5 variants, for token counting.
docs: update environment variable instructions for multi-model support
- Added details for setting `ANTHROPIC_DEFAULT_OPUS_MODEL`, `ANTHROPIC_DEFAULT_SONNET_MODEL`, and `ANTHROPIC_DEFAULT_HAIKU_MODEL` for version 2.x.x.
- Clarified usage of `ANTHROPIC_MODEL` and `ANTHROPIC_SMALL_FAST_MODEL` for version 1.x.x.
- Expanded examples for setting environment variables across different models including Gemini, GPT-5, Claude, and Qwen3.
- Updated the Execute methods in various executors (GeminiCLIExecutor, GeminiExecutor, IFlowExecutor, OpenAICompatExecutor, QwenExecutor) to return a response and error as named return values for improved clarity.
- Enhanced error handling by deferring failure tracking in usage reporters, ensuring that failures are reported correctly.
- Improved response body handling by ensuring proper closure and error logging for HTTP responses across all executors.
- Added failure tracking and reporting in the usage reporter to capture unsuccessful requests.
- Updated the usage logging structure to include a 'Failed' field for better tracking of request outcomes.
- Adjusted the logic in the RequestStatistics and Record methods to accommodate the new failure tracking mechanism.
- Added detailed logging of upstream request metadata including URL, method, headers, and body for Codex, Gemini, IFlow, OpenAI Compat, and Qwen executors.
- Implemented error logging for API response failures to capture errors during HTTP requests.
- Introduced structured logging for authentication details (AuthID, AuthLabel, AuthType, AuthValue) to improve traceability.
- Updated response logging to include status codes and headers for better debugging.
- Ensured that all executors consistently log API interactions to facilitate monitoring and troubleshooting.
- Added new "Gemini 2.5 Flash Image Preview" model definition, with enhanced image generation capabilities.
- Increased scanner buffer size to 20,971,520 bytes across executors and translators to handle larger payloads.
- Refined trimming and normalization logic for `baseURL` and `apiKey` attributes.
- Updated `Authorization` header logic to omit empty API keys.
- Enhanced compatibility processing by handling empty `api-key-entries`.
- Improved legacy format fallback and added safeguards for empty credentials across executor paths.
- Added `newProxyAwareHTTPClient` to centralize proxy configuration with priority on `auth.ProxyURL` and `cfg.ProxyURL`.
- Integrated enhanced proxy support across executors for HTTP, HTTPS, and SOCKS5 protocols.
- Refactored redundant HTTP client initialization to use `newProxyAwareHTTPClient` for consistent behavior.
- Added `LoggerPlugin` to log usage metrics for observability.
- Introduced a new `Manager` to handle usage record queuing and plugin registration.
- Integrated new usage reporter and detailed metrics parsing into executors, covering providers like OpenAI, Codex, Claude, and Gemini.
- Improved token usage breakdown across streaming and non-streaming responses.
- Introduced `CountTokens` method to Codex, Claude, Gemini, Qwen, OpenAI-compatible, and other executors.
- Implemented `ExecuteCount` in `AuthManager` for token counting via provider round-robin.
- Updated handlers to leverage `ExecuteCountWithAuthManager` for streamlined token counting.
- Added fallback and error handling logic for token counting requests.
