- Implemented a Go-based dynamic library plugin for request lifecycle management.
- Added concurrency controls, keyword-based request termination, and response handling.
- Supported optional capabilities for request interception and active lifecycle termination.
- Included tests for schema compatibility, concurrency limits, and policy-based termination.
- Added build instructions and configuration details in README.
- Updated host support for lifecycle plugin RPC methods.
Closes: #4568
- Added detailed specifications for `Gemini 3.1 Pro Preview`, `Gemini 3.5 Flash Lite`, and `Gemini 3.6 Flash` models, including token limits, generation methods, and thought levels.
- Updated `models.json` to include entries for all three versions.
Closes: #4555
- Introduced a new reasoning level, `minimal`, to the `codexClientAllowedReasoningLevels` map.
- Updated response-building logic to include descriptive messaging for the `minimal` level.
Closes: #4542
- Added detailed specifications for `Gemini 3.5 Flash Lite` and `Gemini 3.6 Flash` models, including token limits, generation methods, and thought levels.
- Updated `models.json` to include entries for both versions across relevant segments.
Closes: #4521
- Introduced `DisableCloakingModelList` in `ClaudeCodeConfig` to control model ID cloaking in Anthropic model list responses.
- Updated relevant APIs and handlers to respect the new configuration.
- Added comprehensive tests for enabling/disabling cloaking behavior and config-driven hot reload scenarios.
- Extended example configuration and documentation to include the new setting.
Closes: #4473
- Added handling for `input_image` in `toolOutputContentPart` and updated related parsing logic.
- Introduced `hasToolOutputImagePart` to check content for image-related types.
- Improved `setToolCallOutputContent` to handle structured content with image parts recursively.
- Refined detail extraction and type mapping for `input_image` and `image_url` cases.
Closes: #4458
- Introduced `NormalizeClaudeToolInputSchema` in `util` package to standardize tool input schemas for compatibility with Claude.
- Replaced local `normalizeClaudeToolInputSchema` implementations in translator functions with the new utility method.
- Improved handling of schema validation, union elimination, and property merging.
- Removed redundant code in response and chat-completions translators, streamlining schema normalization logic.
Closes: #4428
- Introduced `InjectXSearch` in `XAIConfig` to enable automatic injection of the native `x_search` tool when not explicitly declared.
- Updated `XAIExecutor` to honor the `InjectXSearch` configuration, ensuring consistent tool availability.
- Enhanced configuration handling with support for dynamic diffing to track changes in `InjectXSearch`.
- Added comprehensive tests to validate `InjectXSearch` behavior, including preparation and tool choice synchronization.
- Updated example config and documentation to outline `InjectXSearch` usage.
Closes: #4339
- Introduced `PostgresCooldownStateStore` for saving and loading cooldown states in a PostgreSQL database.
- Supported runtime cooldown state management with efficient upsert and deletion.
- Added configurable integration via `PostgresStoreConfig` and updated schema management to include cooldown-related tables.
- Included tests to validate persistence, normalization, and concurrent handling of cooldown state records.
- Enabled overriding cooldown state stores via service builder and SDK configuration.
Closes: #4254
- Enhanced handling of unique and ambiguous tool names for `custom` and `function` types.
- Consolidated name validation and shortening logic for clarity.
- Refactored `resolveToolCall` to standardize tool call processing.
- Improved support for tool name disambiguation and metadata mapping.
Closes: #4208
- Added `CountClaudeInputTokens` for estimating token usage with O200kBase tokenizer.
- Enhanced `ClaudeExecutor` to handle token counting locally, reducing dependency on upstream services.
- Introduced validation for token count requests for improved error handling.
- Updated tests to cover local token counting behavior and invalid request handling.
Closes: #4103
- Introduced `toolCallStreamState` to manage tool call state transitions.
- Improved support for `custom_tool_call` types with fallback handling.
- Refactored argument emission logic for better clarity and flexibility.
- Added utility methods for managing and retrieving tool call states.
Closes: #4078
- Introduced `NewLifetime` to create fresh `Client` instances while maintaining cluster failover state.
- Enhanced reconnect failure tracking with `markReconnectFailure` conditions for various connection stages.
- Reset reconnect failure counter on successful reconnections to ensure stable client behavior.
- Updated tests to verify failover mechanics, subscription handling, and heartbeat recovery scenarios.
- Improved lifetime management in `service_home` with reuse of `previousClient` for seamless client transitions.
- Removed deprecated interceptor and executor-related methods, including `callRequestInterceptor`, `callResponseInterceptor`, and `callStreamChunkInterceptor`.
- Consolidated unused logic and pruned redundant imports to streamline `adapters.go`.
- No functional changes.
Keep a single Claude thinking block open across multiple reasoning summary
parts for one Codex reasoning item, and finalize it only when output_item.done
delivers the item's final encrypted_content.
Previously, each summary part closed the preceding thinking block with the
pre-content encrypted_content snapshot captured at output_item.added. This
emitted N thinking blocks with placeholder signatures for one reasoning item,
causing client replay to append redundant placeholder reasoning items into
conversation history.
Add regression tests verifying that multi-part summary reasoning items produce
exactly one thinking block signed with the final encrypted_content.
`prependToFirstUserMessage` always inserted the `<system-reminder>` text block at
index 0 of the first user message's content array. When a client sends a history
that begins with an assistant `tool_use` turn, that first user message is the
`tool_result` carrier, and Anthropic requires those blocks to stay at the head of
the message. Prepending pushed them out of first position, so the upstream
rejected the whole request with:
messages.N: `tool_use` ids were found without `tool_result` blocks
immediately after: <id>
Append the reminder instead when the content array already leads with a
`tool_result` block; behaviour is unchanged for every other message shape.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
- Added logic to `preserveXAIResponsesOutputControls` for retaining supported controls (`max_output_tokens`, `temperature`, `top_p`, `top_k`) during translation.
- Removed unsupported fields such as `stop` where necessary to align with xAI's Responses API.
- Updated tests to validate the preservation and removal of fields across various input scenarios.
- Improved payload configuration handling with additional unit tests for ensuring compatibility with model-specific overrides.
Closes: #4464
- Added `UsesConfig` to `XAIAutoExecutor` for determining config-based binding.
- Refactored `Service` to support thread-safe executor registration, using a new mutex (`executorRegistrationMu`).
- Improved executor rebind logic to prevent unnecessary replacements unless forced or required by config updates.
- Updated handling of config updates to correctly replace stale XAI executors.
- Enhanced the test suite for edge cases in executor binding and replacement for multiple providers.
Closes: #4567
- Updated `ensureCacheControl` to inject cache_control into the last non-deferred tool instead of the last tool.
- Adjusted `injectToolsCacheControl` to ignore tools marked with `defer_loading` for prompt caching.
- Added unit tests to validate behavior for deferred tools, existing cache_control, and mixed tool arrays.
Closes: #4561
Fixes#4514 by sanitizing unsigned thinking signatures for Claude target models in OpenAI Chat Completions requests to Antigravity.
- Export SanitizeAntigravityClaudeGeminiRequestSignatures in gemini request translator and call it for Claude target models in OpenAI Chat Completions requests.
- Use json.Decoder with UseNumber() in SanitizeAntigravityClaudeGeminiRequestSignatures to preserve exact JSON number precision (preventing float64 truncation of large integers and high-precision numbers during unmarshal/marshal).
- Add unit tests verifying unsigned reasoning_content sanitization, empty turn dropping, exact number precision preservation, and functionCall signature stripping for Claude target models.
- Extended `models.json` with Claude Opus 5, a premium high-context model by Anthropic.
- Added Gemini 3.6 Flash (High) by Antigravity, with advanced token and thinking configurations.
- Implemented logging for Codex remote media forwarding start events, including detailed connection and credential metadata.
- Added `formatLogFieldValue` for quoting specific log fields and ensured newline safety in log output.
- Enhanced unit tests to validate log content, escaping, and field inclusion.
- Implemented a TCP proxy for WebRTC candidate tunneling in Codex Live, supporting passive TCP candidates on port 443.
- Restricted tunneling to globally routable public IPs and added safeguards for rejecting unsafe/private targets.
- Added robust validation of STUN BindingRequest frames before forwarding to upstream candidates.
- Includes extensive unit tests for proxying behavior, candidate validation, and tunnel edge cases.
- Added comprehensive diffing for Codex live media relay settings, including support for public IP, UDP port ranges, and ICE server changes.
- Introduced `displayOptionalValue` utility to handle optional values in diff outputs.
- Improved test coverage for config change detection, ensuring no sensitive information leakage.
- Replaced `allow-private-remote-ips` with the new `disable-private-remote-ips` property, adding YAML backward compatibility.
- Updated Codex live handler to differentiate media relay configuration changes and runtime updates.
