Genkit MCP

# This file holds configuration for the genkit-schema.json file # generated by the npm export:schemas script. # ============================================================================ # DOCUMENTATION SECTION # All type and field documentation in one consolidated location # ============================================================================ # ---------------------------------------------------------------------------- # Core Message Types # ---------------------------------------------------------------------------- Role doc Role indicates which entity is responsible for the content of a message. . RoleSystem doc RoleSystem indicates this message is user-independent context. . RoleUser doc RoleUser indicates this message was generated by the client. . RoleModel doc RoleModel indicates this message was generated by the model during a previous interaction. . RoleTool doc RoleTool indicates this message was generated by a local tool, likely triggered by a request from the model in one of its previous responses. . Message doc Message represents the contents of a model message in a conversation. . Message.role doc Role indicates which entity (system, user, model, or tool) generated this message. . Message.content doc Content holds the message parts (text, media, tool calls, etc.). . Message.metadata doc Metadata contains arbitrary key-value data associated with this message. . # ---------------------------------------------------------------------------- # Part Types (Message Content) # ---------------------------------------------------------------------------- TextPart.text doc Text contains the textual content. . TextPart.metadata doc Metadata contains arbitrary key-value data for this part. . MediaPart.media doc Media contains the media content and metadata. . MediaPart.metadata doc Metadata contains arbitrary key-value data for this part. . ToolRequestPart.toolRequest doc ToolRequest is a request for a tool to be executed, usually provided by a model. . ToolRequestPart.metadata doc Metadata contains arbitrary key-value data for this part. . ToolResponsePart.toolResponse doc ToolResponse is a provided response to a tool call. . ToolResponsePart.metadata doc Metadata contains arbitrary key-value data for this part. . DataPart.data doc Data contains arbitrary structured data. . DataPart.metadata doc Metadata contains arbitrary key-value data for this part. . ReasoningPart.reasoning doc Reasoning contains the reasoning text of the message. . ReasoningPart.metadata doc Metadata contains arbitrary key-value data for this part. . CustomPart.custom doc Custom contains custom key-value data specific to this part. . CustomPart.data doc Data contains additional arbitrary data. . CustomPart.metadata doc Metadata contains arbitrary key-value data for this part. . ResourcePart.resource doc Resource contains a reference to an external resource by URI. . ResourcePart.metadata doc Metadata contains arbitrary key-value data for this part. . ResourcePartResource.uri doc Uri is the URI of the external resource. . # ---------------------------------------------------------------------------- # Media Types # ---------------------------------------------------------------------------- Media doc Media represents media content with a URL and content type. . Media.contentType doc ContentType specifies the MIME type of the media. Inferred from the data URI if not provided. . Media.url doc Url is a "data:" or "https:" URI containing the media content. . # ---------------------------------------------------------------------------- # Tool Types # ---------------------------------------------------------------------------- ToolRequest doc A ToolRequest is a message from the model to the client that it should run a specific tool and pass a ToolResponse to the model on the next chat request it makes. Any ToolRequest will correspond to some ToolDefinition previously sent by the client. . ToolRequest.ref doc Ref is the call ID or reference for this specific request. . ToolRequest.name doc Name is the name of the tool to call. . ToolRequest.input doc Input is a JSON object containing the input parameters for the tool. For example: map[string]any{"country":"USA", "president":3}. . ToolRequest.partial doc Partial indicates whether this is a partial streaming chunk. . ToolResponse doc A ToolResponse is a message from the client to the model containing the results of running a specific tool on the arguments passed to the client by the model in a ToolRequest. . ToolResponse.ref doc Ref is the call ID or reference matching the original request. . ToolResponse.name doc Name is the name of the tool that was executed. . ToolResponse.output doc Output is a JSON object describing the results of running the tool. For example: map[string]any{"name":"Thomas Jefferson", "born":1743}. . ToolResponse.content doc Content holds additional message parts that provide context or details about the tool response. . ToolDefinition doc A ToolDefinition describes a tool. . ToolDefinition.name doc Name is the unique identifier for this tool. . ToolDefinition.description doc Description explains what the tool does and when to use it. . ToolDefinition.inputSchema doc InputSchema is a valid JSON Schema representing the input parameters of the tool. . ToolDefinition.outputSchema doc OutputSchema is a valid JSON Schema describing the output of the tool. . ToolDefinition.metadata doc Metadata contains additional information about this tool definition. . # ---------------------------------------------------------------------------- # Generation Configuration # ---------------------------------------------------------------------------- GenerationCommonConfig doc GenerationCommonConfig holds configuration parameters for model generation requests. . GenerationCommonConfig.version doc Version specifies a particular version of a model family, e.g., "gemini-1.0-pro-001" for the "gemini-1.0-pro" family. . GenerationCommonConfig.temperature doc Temperature controls randomness in generation. Higher values (e.g., 0.9) make output more random, while lower values (e.g., 0.1) make it more deterministic. Typical range is 0.0 to 1.0. . GenerationCommonConfig.maxOutputTokens doc MaxOutputTokens limits the maximum number of tokens generated in the response. . GenerationCommonConfig.topK doc TopK limits sampling to the K most likely tokens at each step. . GenerationCommonConfig.topP doc TopP (nucleus sampling) limits sampling to tokens whose cumulative probability exceeds P. . GenerationCommonConfig.stopSequences doc StopSequences specifies sequences that will cause generation to stop when encountered. . # ---------------------------------------------------------------------------- # Generation Usage and Metrics # ---------------------------------------------------------------------------- GenerationUsage doc GenerationUsage provides information about resource consumption during generation. . GenerationUsage.inputTokens doc InputTokens is the number of tokens in the input prompt. . GenerationUsage.outputTokens doc OutputTokens is the number of tokens generated in the response. . GenerationUsage.totalTokens doc TotalTokens is the sum of input and output tokens. . GenerationUsage.inputCharacters doc InputCharacters is the number of characters in the input. . GenerationUsage.outputCharacters doc OutputCharacters is the number of characters generated in the output. . GenerationUsage.inputImages doc InputImages is the number of images in the input. . GenerationUsage.outputImages doc OutputImages is the number of images generated in the output. . GenerationUsage.inputVideos doc InputVideos is the number of videos in the input. . GenerationUsage.outputVideos doc OutputVideos is the number of videos generated in the output. . GenerationUsage.inputAudioFiles doc InputAudioFiles is the number of audio files in the input. . GenerationUsage.outputAudioFiles doc OutputAudioFiles is the number of audio files generated in the output. . GenerationUsage.thoughtsTokens doc ThoughtsTokens counts tokens used in reasoning or thinking processes. . GenerationUsage.cachedContentTokens doc CachedContentTokens counts tokens that were served from cache. . GenerationUsage.custom doc Custom contains additional usage metrics specific to the model provider. . # ---------------------------------------------------------------------------- # Model Request and Response # ---------------------------------------------------------------------------- ModelRequest doc A ModelRequest is a request to generate completions from a model. . ModelRequest.messages doc Messages contains the conversation history for the model. . ModelRequest.config doc Config holds model-specific configuration parameters. . ModelRequest.docs doc Docs provides retrieved documents to be used as context for this generation. . ModelRequest.output doc Output describes the desired response format. . ModelRequest.tools doc Tools lists the available tools that the model can ask the client to run. . ModelRequest.toolChoice doc ToolChoice controls how the model uses tools (auto, required, or none). . ModelResponse doc A ModelResponse is a model's response to a ModelRequest. . ModelResponse.message doc Message contains the generated response content. . ModelResponse.finishReason doc FinishReason indicates why generation stopped (e.g., stop, length, blocked). . ModelResponse.finishMessage doc FinishMessage provides additional details about why generation finished. . ModelResponse.latencyMs doc LatencyMs is the time the request took in milliseconds. . ModelResponse.usage doc Usage describes how many resources were used by this generation request. . ModelResponse.custom doc Custom contains model-specific extra information. Deprecated: use Raw instead. . ModelResponse.raw doc Raw contains the unprocessed model-specific response data. . ModelResponse.request doc Request is the ModelRequest struct used to trigger this response. . ModelResponse.operation doc Operation provides information about a long-running background task if applicable. . ModelResponseChunk doc A ModelResponseChunk is the portion of the ModelResponse that is passed to a streaming callback. . ModelResponseChunk.role doc Role indicates the entity that generated this chunk. . ModelResponseChunk.index doc Index of the message this chunk belongs to. . ModelResponseChunk.content doc Content is the chunk of message parts to stream right now. . ModelResponseChunk.custom doc Custom contains model-specific extra information attached to this chunk. . ModelResponseChunk.aggregated doc Aggregated indicates whether the chunk includes all data from previous chunks. If false, the chunk is considered incremental. . # ---------------------------------------------------------------------------- # Model Information and Capabilities # ---------------------------------------------------------------------------- ModelInfo doc ModelInfo contains metadata about a model's capabilities and characteristics. . ModelInfo.versions doc Versions lists acceptable names for this model (e.g., different versions). . ModelInfo.label doc Label is a friendly display name for this model (e.g., "Google AI - Gemini Pro"). . ModelInfo.configSchema doc ConfigSchema defines the model-specific configuration schema. . ModelInfo.supports doc Supports describes the capabilities that this model supports. . ModelInfo.stage doc Stage indicates the development stage of this model. Featured models are recommended for general use, stable models are well-tested, unstable models are experimental, legacy models are not recommended for new projects, and deprecated models may be removed in future versions. . ModelInfoSupports doc ModelSupports describes the capabilities that a model supports. . ModelInfoSupports.multiturn doc Multiturn indicates whether the model can process historical messages passed with a prompt. . ModelInfoSupports.media doc Media indicates whether the model can process media as part of the prompt (multimodal input). . ModelInfoSupports.tools doc Tools indicates whether the model can perform tool calls. . ModelInfoSupports.systemRole doc SystemRole indicates whether the model can accept messages with role "system". . ModelInfoSupports.output doc Output lists the types of data the model can generate. . ModelInfoSupports.contentType doc ContentType lists the content types the model supports for output. . ModelInfoSupports.context doc Context indicates whether the model can natively support document-based context grounding. . ModelInfoSupports.constrained doc Constrained indicates the level of constrained generation support (none, all, or no-tools). . ModelInfoSupports.toolChoice doc ToolChoice indicates whether the model supports controlling tool choice (e.g., forced tool calling). . ModelInfoSupports.longRunning doc LongRunning indicates whether the model supports long-running operations. . # ---------------------------------------------------------------------------- # Output Configuration # ---------------------------------------------------------------------------- OutputConfig doc OutputConfig describes the structure that the model's output should conform to. If Format is OutputFormatJSON, then Schema can describe the desired form of the generated JSON. . OutputConfig.format doc Format specifies the desired output format (e.g., "json", "text"). . OutputConfig.schema doc Schema is a JSON Schema describing the desired structure of the output. . OutputConfig.constrained doc Constrained indicates whether to enforce strict adherence to the schema. . OutputConfig.contentType doc ContentType specifies the MIME type of the output content. . # ---------------------------------------------------------------------------- # Operation Types # ---------------------------------------------------------------------------- Operation doc Operation represents a long-running background task. . Operation.action doc Action is the name of the action being performed by this operation. . Operation.id doc Id is the unique identifier for this operation. . Operation.done doc Done indicates whether the operation has completed. . Operation.output doc Output contains the result of the operation if it has completed successfully. . Operation.error doc Error contains error information if the operation failed. . Operation.metadata doc Metadata contains additional information about the operation. . OperationError doc OperationError contains error information for a failed operation. . OperationError.message doc Message describes the error that occurred. . # ---------------------------------------------------------------------------- # Document Types # ---------------------------------------------------------------------------- # Note: Document type is hand-written in ai/document.go, not generated # ---------------------------------------------------------------------------- # Embedding Types # ---------------------------------------------------------------------------- Embedding doc Embedding represents a vector embedding with associated metadata. . Embedding.embedding doc Embedding is the vector representation of the input. . Embedding.metadata doc Metadata identifies which part of a document this embedding corresponds to. . EmbedRequest doc EmbedRequest represents a request to generate embeddings for documents. . EmbedRequest.input doc Input is the array of documents to generate embeddings for. . EmbedRequest.options doc Options contains embedder-specific configuration parameters. . EmbedResponse doc EmbedResponse contains the generated embeddings from an embed request. . EmbedResponse.embeddings doc Embeddings is the array of generated embedding vectors with metadata. . # ---------------------------------------------------------------------------- # Evaluator Types (ScoreDetails only - other eval types are omitted) # ---------------------------------------------------------------------------- ScoreDetails doc ScoreDetails provides additional context and explanation for an evaluation score. . ScoreDetails.reasoning doc Reasoning explains the rationale behind the score. . # ---------------------------------------------------------------------------- # Retriever Types # ---------------------------------------------------------------------------- RetrieverRequest doc RetrieverRequest represents a request to retrieve relevant documents. . RetrieverRequest.query doc Query is the document to use for retrieval. . RetrieverRequest.options doc Options contains retriever-specific configuration parameters. . RetrieverResponse doc RetrieverResponse contains the retrieved documents from a retriever request. . RetrieverResponse.documents doc Documents is the array of retrieved documents. . # ---------------------------------------------------------------------------- # Reranker Types # ---------------------------------------------------------------------------- RerankerRequest doc RerankerRequest represents a request to rerank documents based on relevance. . RerankerRequest.query doc Query is the document to use for reranking. . RerankerRequest.documents doc Documents is the array of documents to rerank. . RerankerRequest.options doc Options contains reranker-specific configuration parameters. . RerankerResponse doc RerankerResponse contains the reranked documents with relevance scores. . RerankerResponse.documents doc Documents is the array of reranked documents with scores. . RankedDocumentData doc RankedDocumentData represents a document with a relevance score from reranking. . RankedDocumentData.content doc Content holds the document's parts (text and media). . RankedDocumentData.metadata doc Metadata contains the reranking score and other arbitrary key-value data. . RankedDocumentMetadata doc RankedDocumentMetadata contains the relevance score and other metadata for a reranked document. . RankedDocumentMetadata.score doc Score is the relevance score assigned by the reranker. . # ---------------------------------------------------------------------------- # GenerateAction Types # ---------------------------------------------------------------------------- GenerateActionOptions doc GenerateActionOptions holds configuration for a generate action request. . GenerateActionOptions.model doc Model is a model name (e.g., "vertexai/gemini-1.0-pro"). . GenerateActionOptions.docs doc Docs provides retrieved documents to be used as context for this generation. . GenerateActionOptions.messages doc Messages contains the conversation history for multi-turn prompting when supported. . GenerateActionOptions.tools doc Tools is a list of registered tool names for this generation if supported. . GenerateActionOptions.toolChoice doc ToolChoice controls tool calling mode. Auto lets the model decide, required forces the model to choose a tool, and none forces the model not to use any tools. Defaults to auto. . GenerateActionOptions.config doc Config contains configuration parameters for the generation request. . GenerateActionOptions.output doc Output specifies the desired output format. Defaults to the model's default if unspecified. . GenerateActionOptions.resume doc Resume provides options for resuming an interrupted generation. . GenerateActionOptions.returnToolRequests doc ReturnToolRequests, when true, returns tool calls for manual processing instead of automatically resolving them. . GenerateActionOptions.maxTurns doc MaxTurns is the maximum number of tool call iterations that can be performed in a single generate call. Defaults to 5. . GenerateActionOptions.stepName doc StepName is a custom step name for this generate call to display in trace views. Defaults to "generate". . GenerateActionOptionsResume doc GenerateActionResume holds options for resuming an interrupted generation. . GenerateActionOptionsResume.respond doc Respond contains tool response parts to send to the model when resuming. . GenerateActionOptionsResume.restart doc Restart contains tool request parts to restart when resuming. . GenerateActionOptionsResume.metadata doc Metadata contains additional context for resuming the generation. . GenerateActionOutputConfig doc GenerateActionOutputConfig specifies the desired output format for a generate action. . GenerateActionOutputConfig.format doc Format specifies the desired output format (e.g., "json", "text"). . GenerateActionOutputConfig.contentType doc ContentType specifies the MIME type of the output content. . GenerateActionOutputConfig.instructions doc Instructions provides additional guidance for the output format. . GenerateActionOutputConfig.jsonSchema doc JsonSchema is a JSON Schema describing the desired structure of JSON output. . GenerateActionOutputConfig.constrained doc Constrained indicates whether to enforce strict adherence to the schema. . GenerateActionOptionsToolChoice doc ToolChoice controls how the model uses tools. . # ---------------------------------------------------------------------------- # Finish Reason Enum # ---------------------------------------------------------------------------- FinishReason doc FinishReason indicates why generation stopped. . # ---------------------------------------------------------------------------- # Model Stage Enum # ---------------------------------------------------------------------------- ModelInfoStage doc ModelStage indicates the development stage of a model. . # ---------------------------------------------------------------------------- # Constrained Support Enum # ---------------------------------------------------------------------------- ModelInfoSupportsConstrained doc ConstrainedSupport indicates the level of constrained generation support. . # ---------------------------------------------------------------------------- # Trace Metadata Types # ---------------------------------------------------------------------------- TraceMetadata doc TraceMetadata contains metadata about a trace execution. . TraceMetadata.featureName doc FeatureName identifies the feature being traced. . TraceMetadata.paths doc Paths contains metadata for each path executed during the trace. . TraceMetadata.timestamp doc Timestamp is when the trace was created. . PathMetadata doc PathMetadata contains metadata about a single execution path in a trace. . PathMetadata.path doc Path is the identifier for this execution path. . PathMetadata.status doc Status indicates the outcome of this path. . PathMetadata.latency doc Latency is the execution time for this path in milliseconds. . PathMetadata.error doc Error contains error information if the path failed. . # ---------------------------------------------------------------------------- # Multipart Tool Response # ---------------------------------------------------------------------------- MultipartToolResponse doc MultipartToolResponse represents a tool response with both structured output and content parts. . MultipartToolResponse.output doc Output contains the structured output data from the tool. . MultipartToolResponse.content doc Content holds additional message parts providing context or details. . # ============================================================================ # CONFIGURATION SECTION # Type mappings, omissions, and other non-documentation directives # ============================================================================ # DocumentData type was hand-written. DocumentData omit # Map DocumentData to Document DocumentData name Document # Code generator cannot handle "allOf" schemas. Part omit # Map DocumentPart to Part DocumentPart name Part # Tracing types were written manually. InstrumentationLibrary omit Link omit SpanContext omit SpanData omit SpanDataSameProcessAsParentSpan omit SpanDataTimeEvents omit SpanMetadata omit SpanMetadataState omit SpanStatus omit TimeEvent omit TimeEventAnnotation omit TraceData omit SpanStartEvent omit SpanEndEvent omit # Typo in schema definition... SpantEventBase omit TraceEvent omit GenerationCommonConfig.maxOutputTokens type int GenerationCommonConfig.topK type int # Unused evaluation types BaseDataPoint omit BaseEvalDataPoint omit EvalFnResponse omit EvalRequest omit EvalResponse omit EvalStatusEnum omit # Unused error types CandidateError omit CandidateErrorCode omit Candidate omit # Unused retriever/reranker option types CommonRerankerOptions omit CommonRetrieverOptions omit DocumentData pkg ai GenerateResponse omit GenerateResponseChunk omit GenerateRequest omit GenerateRequestToolChoice omit GenerationUsage pkg ai GenerationUsage.inputCharacters type int GenerationUsage.inputImages type int GenerationUsage.inputAudioFiles type int GenerationUsage.inputVideos type int GenerationUsage.inputTokens type int GenerationUsage.outputCharacters type int GenerationUsage.outputImages type int GenerationUsage.outputAudioFiles type int GenerationUsage.outputVideos type int GenerationUsage.outputTokens type int GenerationUsage.totalTokens type int GenerationUsage.thoughtsTokens type int GenerationUsage.cachedContentTokens type int GenerationCommonConfig pkg ai Message pkg ai ToolDefinition pkg ai ToolDefinition.inputSchema type map[string]any ToolDefinition.outputSchema type map[string]any Part pkg ai TextPart pkg ai TextPart name textPart TextPart.data omit TextPart.media omit TextPart.toolRequest omit TextPart.toolResponse omit TextPart.custom omit TextPart.metadata type map[string]any TextPart.reasoning omit TextPart.resource omit MediaPart pkg ai MediaPart name mediaPart MediaPart.text omit MediaPart.toolRequest omit MediaPart.toolResponse omit MediaPart.custom omit MediaPart.data omit MediaPart.metadata type map[string]any MediaPart.reasoning omit MediaPart.resource omit ToolRequestPart pkg ai ToolRequestPart name toolRequestPart ToolRequestPart.text omit ToolRequestPart.media omit ToolRequestPart.data omit ToolRequestPart.toolResponse omit ToolRequestPart.custom omit ToolRequestPart.metadata type map[string]any ToolRequestPart.reasoning omit ToolRequestPart.resource omit ToolResponsePart pkg ai ToolResponsePart name toolResponsePart ToolResponsePart.text omit ToolResponsePart.media omit ToolResponsePart.data omit ToolResponsePart.toolRequest omit ToolResponsePart.custom omit ToolResponsePart.metadata type map[string]any ToolResponsePart.reasoning omit ToolResponsePart.resource omit ToolResponse.content type []*Part DataPart pkg ai DataPart name dataPart DataPart.media omit DataPart.text omit DataPart.toolRequest omit DataPart.toolResponse omit DataPart.custom omit DataPart.metadata type map[string]any DataPart.reasoning omit DataPart.resource omit ReasoningPart pkg ai ReasoningPart name reasoningPart ReasoningPart.media omit ReasoningPart.data omit ReasoningPart.text omit ReasoningPart.toolRequest omit ReasoningPart.toolResponse omit ReasoningPart.custom omit ReasoningPart.metadata type map[string]any ReasoningPart.resource omit CustomPart pkg ai CustomPart name customPart ResourcePart pkg ai ResourcePart name resourcePart ResourcePart.text omit ResourcePart.media omit ResourcePart.toolRequest omit ResourcePart.toolResponse omit ResourcePart.custom omit ResourcePart.data omit ResourcePart.metadata type map[string]any ResourcePart.reasoning omit ResourcePart.resource type *ResourcePart # Use the auto-generated ResourcePartResource type for consistency with schema ResourcePartResource pkg ai ResourcePartResource name ResourcePart ModelInfo pkg ai ModelInfoSupports pkg ai ModelInfoSupports.output type []string ModelInfoSupports name ModelSupports ModelInfoSupportsConstrained name ConstrainedSupport ModelInfoStage name ModelStage Role pkg ai RoleUser pkg ai RoleModel pkg ai RoleTool pkg ai # GenerateActionOptions GenerateActionOptions pkg ai GenerateActionOptions.model type string GenerateActionOptions.docs type []*Document GenerateActionOptions.messages type []*Message GenerateActionOptions.tools type []string GenerateActionOptionsToolChoice name ToolChoice GenerateActionOptions.config type any GenerateActionOptions.output type *GenerateActionOutputConfig GenerateActionOptions.returnToolRequests type bool GenerateActionOptions.maxTurns type int GenerateActionOptionsResume name GenerateActionResume # GenerateActionOutputConfig GenerateActionOutputConfig.instructions type *string GenerateActionOutputConfig.format name OutputConfigFormat GenerateActionOutputConfig.jsonSchema name Schema GenerateActionOutputConfig.jsonSchema type map[string]any GenerateActionOutputConfig.constrained type bool # ModelRequest ModelRequest pkg ai ModelRequest.config type any ModelRequest.messages type []*Message ModelRequest.output type *ModelOutputConfig ModelRequest.tools type []*ToolDefinition ModelRequest.toolChoice type ToolChoice ModelRequest.docs type []*Document OutputConfig name ModelOutputConfig # ModelResponse ModelResponse pkg ai ModelResponse.custom type any ModelResponse.finishMessage type string ModelResponse.latencyMs type float64 ModelResponse.message type *Message ModelResponse.request type *ModelRequest ModelResponse.usage type *GenerationUsage ModelResponse.raw type any ModelResponse.operation type *Operation ModelResponse field formatHandler StreamingFormatHandler # ModelResponseChunk ModelResponseChunk pkg ai ModelResponseChunk.aggregated type bool ModelResponseChunk.content type []*Part ModelResponseChunk.custom type any ModelResponseChunk.index type int ModelResponseChunk.role type Role ModelResponseChunk field formatHandler StreamingFormatHandler Score omit Embedding.embedding type []float32 GenkitError omit GenkitErrorData omit GenkitErrorDataGenkitErrorDetails omit