类的使用
io.agentscope.core.message.Msg

使用Msg的程序包

程序包	说明
io.agentscope.core
io.agentscope.core.agent
io.agentscope.core.agent.user
io.agentscope.core.formatter
io.agentscope.core.formatter.anthropic
io.agentscope.core.formatter.dashscope
io.agentscope.core.formatter.gemini
io.agentscope.core.formatter.ollama
io.agentscope.core.formatter.openai
io.agentscope.core.hook
io.agentscope.core.interruption
io.agentscope.core.memory
io.agentscope.core.message
io.agentscope.core.model
io.agentscope.core.pipeline
io.agentscope.core.plan
io.agentscope.core.rag.model
io.agentscope.core.tool
io.agentscope.core.tracing
io.agentscope.core.util

io.agentscope.core中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core中的方法

修饰符和类型	方法	说明
protected reactor.core.publisher.Mono<Msg>	ReActAgent.doCall(List<Msg> msgs)
protected reactor.core.publisher.Mono<Msg>	ReActAgent.handleInterrupt(InterruptContext context, Msg... originalArgs)
protected reactor.core.publisher.Mono<Msg>	ReActAgent.summarizing()	Generate summary when max iterations reached.

参数类型为Msg的io.agentscope.core中的方法

修饰符和类型	方法	说明
protected reactor.core.publisher.Mono<Void>	ReActAgent.doObserve(Msg msg)
protected reactor.core.publisher.Mono<Msg>	ReActAgent.handleInterrupt(InterruptContext context, Msg... originalArgs)

类型变量类型为Msg的io.agentscope.core中的方法参数

修饰符和类型	方法	说明
protected reactor.core.publisher.Mono<Msg>	ReActAgent.doCall(List<Msg> msgs)

io.agentscope.core.agent中Msg的使用

返回Msg的io.agentscope.core.agent中的方法

修饰符和类型	方法	说明
Msg	Event.getMessage()	Get the message content.
Msg	StructuredOutputHook.getResultMsg()	Get the result message from generate_response.

返回变量类型为Msg的类型的io.agentscope.core.agent中的方法

修饰符和类型	方法	说明
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs)	Process a list of input messages and generate a response with hook execution.
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode schema)	Process multiple input messages and generate structured output with hook execution.
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs, Class<?> structuredOutputClass)	Process multiple input messages and generate structured output with hook execution.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call()	Continue generation based on current state without adding new input.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(com.fasterxml.jackson.databind.JsonNode schema)	Continue generation with JSON schema based on current state.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg)	Process a single input message and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg... msgs)	Process multiple input messages (varargs) and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg, com.fasterxml.jackson.databind.JsonNode schema)	Process a single input message with JSON schema and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg, Class<?> structuredModel)	Process a single input message with structured model and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Class<?> structuredModel)	Continue generation with structured model based on current state.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs)	Process a list of input messages and generate a response.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode schema)	Process multiple input messages with JSON schema and generate a response.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs, Class<?> structuredModel)	Process multiple input messages with structured model and generate a response.
protected abstract reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs)	Internal implementation for processing multiple input messages.
protected reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode outputSchema)	Internal implementation for processing multiple messages with structured output.
protected reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs, Class<?> structuredOutputClass)	Internal implementation for processing multiple messages with structured output.
protected final reactor.core.publisher.Mono<Msg>	StructuredOutputCapableAgent.doCall(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode outputSchema)
protected final reactor.core.publisher.Mono<Msg>	StructuredOutputCapableAgent.doCall(List<Msg> msgs, Class<?> structuredOutputClass)
protected abstract reactor.core.publisher.Mono<Msg>	AgentBase.handleInterrupt(InterruptContext context, Msg... originalArgs)	Handle an interruption that occurred during execution.

参数类型为Msg的io.agentscope.core.agent中的方法

修饰符和类型	方法	说明
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg)	Process a single input message and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg... msgs)	Process multiple input messages (varargs) and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg, com.fasterxml.jackson.databind.JsonNode schema)	Process a single input message with JSON schema and generate a response.
default reactor.core.publisher.Mono<Msg>	CallableAgent.call(Msg msg, Class<?> structuredModel)	Process a single input message with structured model and generate a response.
protected reactor.core.publisher.Mono<Void>	AgentBase.doObserve(Msg msg)	Observe a message without generating a reply.
protected abstract reactor.core.publisher.Mono<Msg>	AgentBase.handleInterrupt(InterruptContext context, Msg... originalArgs)	Handle an interruption that occurred during execution.
void	Agent.interrupt(Msg msg)	Interrupt the current agent execution with a user message.
void	AgentBase.interrupt(Msg msg)	Interrupt the current agent execution with a user message.
final reactor.core.publisher.Mono<Void>	AgentBase.observe(Msg msg)	Observe a single message without generating a reply.
reactor.core.publisher.Mono<Void>	ObservableAgent.observe(Msg msg)	Observe a single message without generating a reply.
default reactor.core.publisher.Flux<Event>	StreamableAgent.stream(Msg msg)	Stream execution events for a single message with default options.
default reactor.core.publisher.Flux<Event>	StreamableAgent.stream(Msg msg, StreamOptions options)	Stream execution events for a single message.
default reactor.core.publisher.Flux<Event>	StreamableAgent.stream(Msg msg, StreamOptions options, com.fasterxml.jackson.databind.JsonNode schema)	Stream execution events for a single message with JSON schema support.
default reactor.core.publisher.Flux<Event>	StreamableAgent.stream(Msg msg, StreamOptions options, Class<?> structuredModel)	Stream execution events for a single message with structured output support.

类型变量类型为Msg的io.agentscope.core.agent中的方法参数

修饰符和类型	方法	说明
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs)	Process a list of input messages and generate a response with hook execution.
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode schema)	Process multiple input messages and generate structured output with hook execution.
final reactor.core.publisher.Mono<Msg>	AgentBase.call(List<Msg> msgs, Class<?> structuredOutputClass)	Process multiple input messages and generate structured output with hook execution.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs)	Process a list of input messages and generate a response.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode schema)	Process multiple input messages with JSON schema and generate a response.
reactor.core.publisher.Mono<Msg>	CallableAgent.call(List<Msg> msgs, Class<?> structuredModel)	Process multiple input messages with structured model and generate a response.
protected abstract reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs)	Internal implementation for processing multiple input messages.
protected reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode outputSchema)	Internal implementation for processing multiple messages with structured output.
protected reactor.core.publisher.Mono<Msg>	AgentBase.doCall(List<Msg> msgs, Class<?> structuredOutputClass)	Internal implementation for processing multiple messages with structured output.
protected final reactor.core.publisher.Mono<Msg>	StructuredOutputCapableAgent.doCall(List<Msg> msgs, com.fasterxml.jackson.databind.JsonNode outputSchema)
protected final reactor.core.publisher.Mono<Msg>	StructuredOutputCapableAgent.doCall(List<Msg> msgs, Class<?> structuredOutputClass)
final reactor.core.publisher.Mono<Void>	AgentBase.observe(List<Msg> msgs)	Observe multiple messages without generating a reply.
reactor.core.publisher.Mono<Void>	ObservableAgent.observe(List<Msg> msgs)	Observe multiple messages without generating a reply.
final reactor.core.publisher.Flux<Event>	AgentBase.stream(List<Msg> msgs, StreamOptions options)	Stream with multiple input messages.
final reactor.core.publisher.Flux<Event>	AgentBase.stream(List<Msg> msgs, StreamOptions options, com.fasterxml.jackson.databind.JsonNode schema)	Stream with multiple input messages using a JSON schema.
final reactor.core.publisher.Flux<Event>	AgentBase.stream(List<Msg> msgs, StreamOptions options, Class<?> structuredModel)	Stream with multiple input messages.
default reactor.core.publisher.Flux<Event>	StreamableAgent.stream(List<Msg> msgs)	Stream execution events for multiple messages with default options.
reactor.core.publisher.Flux<Event>	StreamableAgent.stream(List<Msg> msgs, StreamOptions options)	Stream execution events in real-time as the agent processes the input.
reactor.core.publisher.Flux<Event>	StreamableAgent.stream(List<Msg> msgs, StreamOptions options, com.fasterxml.jackson.databind.JsonNode schema)	Stream execution events with JSON schema support.
reactor.core.publisher.Flux<Event>	StreamableAgent.stream(List<Msg> msgs, StreamOptions options, Class<?> structuredModel)	Stream execution events with structured output support.

参数类型为Msg的io.agentscope.core.agent中的构造器

限定符	构造器	说明
	Event(EventType type, Msg message, boolean isLast)	Creates a new event.

io.agentscope.core.agent.user中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.agent.user中的方法

修饰符和类型	方法	说明
protected reactor.core.publisher.Mono<Msg>	UserAgent.handleInterrupt(InterruptContext context, Msg... originalArgs)	Handle interrupt scenarios.

参数类型为Msg的io.agentscope.core.agent.user中的方法

修饰符和类型	方法	说明
protected reactor.core.publisher.Mono<Msg>	UserAgent.handleInterrupt(InterruptContext context, Msg... originalArgs)	Handle interrupt scenarios.

类型变量类型为Msg的io.agentscope.core.agent.user中的方法参数

修饰符和类型	方法	说明
reactor.core.publisher.Mono<UserInputData>	StreamUserInput.handleInput(String agentId, String agentName, List<Msg> contextMessages, Class<?> structuredModel)	Handle user input from the configured input stream.
reactor.core.publisher.Mono<UserInputData>	UserInputBase.handleInput(String agentId, String agentName, List<Msg> contextMessages, Class<?> structuredModel)	Handle user input and return the input data.

io.agentscope.core.formatter中Msg的使用

参数类型为Msg的io.agentscope.core.formatter中的方法

修饰符和类型	方法	说明
protected String	AbstractBaseFormatter.extractTextContent(Msg msg)	Extract text content from a message, filtering out ThinkingBlock.
protected boolean	AbstractBaseFormatter.hasMediaContent(Msg msg)	Check if a message contains multimodal content (images, audio, video).
protected boolean	AbstractBaseFormatter.shouldBypassHistory(Msg msg)	Check if a message should bypass history merging in multiagent formatters.

类型变量类型为Msg的io.agentscope.core.formatter中的方法参数

修饰符和类型	方法	说明
protected abstract List<TReq>	AbstractBaseFormatter.doFormat(List<Msg> msgs)
List<TReq>	AbstractBaseFormatter.format(List<Msg> msgs)	Format AgentScope messages to provider-specific request format.
List<TReq>	Formatter.format(List<Msg> msgs)	Format AgentScope messages to provider-specific request format.

io.agentscope.core.formatter.anthropic中Msg的使用

类型变量类型为Msg的io.agentscope.core.formatter.anthropic中的方法参数

修饰符和类型	方法	说明
void	AnthropicBaseFormatter.applySystemMessage(com.anthropic.models.messages.MessageCreateParams.Builder paramsBuilder, List<Msg> messages)	Extract and apply system message if present.
List<com.anthropic.models.messages.MessageParam>	AnthropicMessageConverter.convert(List<Msg> messages)	Convert list of Msg to list of Anthropic MessageParam.
List<com.anthropic.models.messages.MessageParam>	AnthropicChatFormatter.doFormat(List<Msg> msgs)
List<com.anthropic.models.messages.MessageParam>	AnthropicMultiAgentFormatter.doFormat(List<Msg> msgs)
String	AnthropicMessageConverter.extractSystemMessage(List<Msg> messages)	Extract system message content if present in the first message.
static List<Object>	AnthropicConversationMerger.mergeConversation(List<Msg> messages, String conversationHistoryPrompt)	Merge conversation messages into formatted blocks with history tags.

io.agentscope.core.formatter.dashscope中Msg的使用

参数类型为Msg的io.agentscope.core.formatter.dashscope中的方法

修饰符和类型	方法	说明
DashScopeMessage	DashScopeMessageConverter.convertToMessage(Msg msg, boolean useMultimodalFormat)	Convert single Msg to DashScopeMessage.

类型变量类型为Msg的io.agentscope.core.formatter.dashscope中的方法参数

修饰符和类型	方法	说明
protected List<DashScopeMessage>	DashScopeChatFormatter.doFormat(List<Msg> msgs)
protected List<DashScopeMessage>	DashScopeMultiAgentFormatter.doFormat(List<Msg> msgs)
List<DashScopeMessage>	DashScopeChatFormatter.formatMultiModal(List<Msg> messages)	Format AgentScope Msg objects to DashScope MultiModal message format.
List<DashScopeMessage>	DashScopeMultiAgentFormatter.formatMultiModal(List<Msg> msgs)	Format AgentScope Msg objects to DashScope MultiModal message format.
DashScopeMessage	DashScopeConversationMerger.mergeToMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)	Merge conversation messages into a single DashScopeMessage (text-only mode).
DashScopeMessage	DashScopeConversationMerger.mergeToMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)	Merge conversation messages into a single DashScopeMessage (text-only mode).
DashScopeMessage	DashScopeConversationMerger.mergeToMultiModalMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, boolean isFirst)	Merge conversation messages into a single DashScopeMessage (multimodal mode).
DashScopeMessage	DashScopeConversationMerger.mergeToMultiModalMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, boolean isFirst)	Merge conversation messages into a single DashScopeMessage (multimodal mode).

io.agentscope.core.formatter.gemini中Msg的使用

类型变量类型为Msg的io.agentscope.core.formatter.gemini中的方法参数

修饰符和类型	方法	说明
List<com.google.genai.types.Content>	GeminiMessageConverter.convertMessages(List<Msg> msgs)	Convert a list of Msg objects to Gemini API Content objects.
protected List<com.google.genai.types.Content>	GeminiChatFormatter.doFormat(List<Msg> msgs)
protected List<com.google.genai.types.Content>	GeminiMultiAgentFormatter.doFormat(List<Msg> msgs)
com.google.genai.types.Content	GeminiConversationMerger.mergeToContent(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)	Merge conversation messages into a single Content (for Gemini API).
com.google.genai.types.Content	GeminiConversationMerger.mergeToContent(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)	Merge conversation messages into a single Content (for Gemini API).

io.agentscope.core.formatter.ollama中Msg的使用

参数类型为Msg的io.agentscope.core.formatter.ollama中的方法

修饰符和类型	方法	说明
OllamaMessage	OllamaMessageConverter.convertMessage(Msg msg)	Converts an AgentScope Msg object to an OllamaMessage object.

类型变量类型为Msg的io.agentscope.core.formatter.ollama中的方法参数

修饰符和类型	方法	说明
protected List<OllamaMessage>	OllamaChatFormatter.doFormat(List<Msg> msgs)
protected List<OllamaMessage>	OllamaMultiAgentFormatter.doFormat(List<Msg> msgs)
OllamaMessage	OllamaConversationMerger.mergeToMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)
OllamaMessage	OllamaConversationMerger.mergeToMessage(List<Msg> msgs, Function<Msg,String> nameExtractor, Function<List<ContentBlock>,String> toolResultConverter, String historyPrompt)

io.agentscope.core.formatter.openai中Msg的使用

参数类型为Msg的io.agentscope.core.formatter.openai中的方法

修饰符和类型	方法	说明
OpenAIMessage	OpenAIMessageConverter.convertToMessage(Msg msg, boolean hasMediaContent)	Convert single Msg to OpenAI DTO Message.

类型变量类型为Msg的io.agentscope.core.formatter.openai中的方法参数

修饰符和类型	方法	说明
protected List<OpenAIMessage>	DeepSeekFormatter.doFormat(List<Msg> msgs)
protected List<OpenAIMessage>	DeepSeekMultiAgentFormatter.doFormat(List<Msg> msgs)
protected List<OpenAIMessage>	GLMFormatter.doFormat(List<Msg> msgs)
protected List<OpenAIMessage>	GLMMultiAgentFormatter.doFormat(List<Msg> msgs)
protected List<OpenAIMessage>	OpenAIChatFormatter.doFormat(List<Msg> msgs)
protected List<OpenAIMessage>	OpenAIMultiAgentFormatter.doFormat(List<Msg> msgs)
OpenAIMessage	OpenAIConversationMerger.mergeToUserMessage(List<Msg> msgs, Function<Msg,String> roleFormatter, Function<List<ContentBlock>,String> toolResultConverter)	Merge conversation messages into a single OpenAIMessage.
OpenAIMessage	OpenAIConversationMerger.mergeToUserMessage(List<Msg> msgs, Function<Msg,String> roleFormatter, Function<List<ContentBlock>,String> toolResultConverter)	Merge conversation messages into a single OpenAIMessage.

类型变量类型为Msg的io.agentscope.core.formatter.openai中的构造器参数

限定符	构造器	说明
	OpenAIMessageConverter(Function<Msg,String> textExtractor, Function<List<ContentBlock>,String> toolResultConverter)	Create an OpenAIMessageConverter with required dependency functions.

io.agentscope.core.hook中Msg的使用

返回Msg的io.agentscope.core.hook中的方法

修饰符和类型	方法	说明
Msg	ReasoningChunkEvent.getAccumulated()	Get the full accumulated message containing all content generated so far.
Msg	SummaryChunkEvent.getAccumulated()	Get the full accumulated message containing all content generated so far.
Msg	PostCallEvent.getFinalMessage()	Get the final response message returned by the agent.
Msg	ReasoningChunkEvent.getIncrementalChunk()	Get only the new content generated in this streaming event.
Msg	SummaryChunkEvent.getIncrementalChunk()	Get only the new content generated in this streaming event.
Msg	PostReasoningEvent.getReasoningMessage()	Get the reasoning result message from LLM.
Msg	PostSummaryEvent.getSummaryMessage()	Get the summary result message.
Msg	PostActingEvent.getToolResultMsg()	Get the tool result message.

返回变量类型为Msg的类型的io.agentscope.core.hook中的方法

修饰符和类型	方法	说明
List<Msg>	PostReasoningEvent.getGotoReasoningMsgs()	Get the messages to add before going back to reasoning.
List<Msg>	PreCallEvent.getInputMessages()	Get the input messages for the agent call.
List<Msg>	PreReasoningEvent.getInputMessages()	Get the messages that will be sent to LLM for reasoning.
List<Msg>	PreSummaryEvent.getInputMessages()	Get the messages that will be sent to LLM for summary generation.

参数类型为Msg的io.agentscope.core.hook中的方法

修饰符和类型	方法	说明
void	PostReasoningEvent.gotoReasoning(Msg msg)	Request to go back to reasoning phase with a single message.
void	PostCallEvent.setFinalMessage(Msg finalMessage)	Modify the final response message.
void	PostReasoningEvent.setReasoningMessage(Msg reasoningMessage)	Modify the reasoning result message.
void	PostSummaryEvent.setSummaryMessage(Msg summaryMessage)	Modify the summary result message.
void	PostActingEvent.setToolResultMsg(Msg toolResultMsg)	Set the tool result message.

类型变量类型为Msg的io.agentscope.core.hook中的方法参数

修饰符和类型	方法	说明
void	PostReasoningEvent.gotoReasoning(List<Msg> msgs)	Request to go back to reasoning phase with multiple messages.
void	PreCallEvent.setInputMessages(List<Msg> inputMessages)	Modify the input messages for the agent call.
void	PreReasoningEvent.setInputMessages(List<Msg> inputMessages)	Modify the messages to send to LLM.
void	PreSummaryEvent.setInputMessages(List<Msg> inputMessages)	Modify the messages to send to LLM for summary.

参数类型为Msg的io.agentscope.core.hook中的构造器

限定符	构造器	说明
	PostCallEvent(Agent agent, Msg finalMessage)	Constructor for PostCallEvent.
	PostReasoningEvent(Agent agent, String modelName, GenerateOptions generateOptions, Msg reasoningMessage)	Constructor for PostReasoningEvent.
	PostSummaryEvent(Agent agent, String modelName, GenerateOptions generateOptions, Msg summaryMessage)	Constructor for PostSummaryEvent.
	ReasoningChunkEvent(Agent agent, String modelName, GenerateOptions generateOptions, Msg incrementalChunk, Msg accumulated)	Constructor for ReasoningChunkEvent.
	SummaryChunkEvent(Agent agent, String modelName, GenerateOptions generateOptions, Msg incrementalChunk, Msg accumulated)	Constructor for SummaryChunkEvent.

类型变量类型为Msg的io.agentscope.core.hook中的构造器参数

限定符	构造器	说明
	PreCallEvent(Agent agent, List<Msg> inputMessages)	Constructor for PreCallEvent.
	PreReasoningEvent(Agent agent, String modelName, GenerateOptions generateOptions, List<Msg> inputMessages)	Constructor for PreReasoningEvent.
	PreSummaryEvent(Agent agent, String modelName, GenerateOptions generateOptions, List<Msg> inputMessages, int maxIterations, int currentIteration)	Constructor for PreSummaryEvent.

io.agentscope.core.interruption中Msg的使用

返回Msg的io.agentscope.core.interruption中的方法

修饰符和类型	方法	说明
Msg	InterruptContext.getUserMessage()	Get the user message that triggered the interruption (if any).

参数类型为Msg的io.agentscope.core.interruption中的方法

修饰符和类型	方法	说明
InterruptContext.Builder	InterruptContext.Builder.userMessage(Msg userMessage)	Set the user message that triggered the interruption.

io.agentscope.core.memory中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.memory中的方法

修饰符和类型	方法	说明
List<Msg>	InMemoryMemory.getMessages()	Retrieves all non-null messages from memory.
List<Msg>	Memory.getMessages()	Retrieves all messages stored in memory.

参数类型为Msg的io.agentscope.core.memory中的方法

修饰符和类型	方法	说明
void	InMemoryMemory.addMessage(Msg message)	Adds a message to the in-memory message list.
void	Memory.addMessage(Msg message)	Adds a message to the memory.
reactor.core.publisher.Mono<String>	LongTermMemory.retrieve(Msg msg)	Retrieves relevant information from long-term memory based on the input message.

类型变量类型为Msg的io.agentscope.core.memory中的方法参数

修饰符和类型	方法	说明
reactor.core.publisher.Mono<Void>	LongTermMemory.record(List<Msg> msgs)	Records messages to long-term memory.

io.agentscope.core.message中Msg的使用

返回Msg的io.agentscope.core.message中的方法

修饰符和类型	方法	说明
Msg	Msg.Builder.build()	Builds a new message instance with the configured properties.
Msg	Msg.withGenerateReason(GenerateReason reason)	Creates a new message with the specified generate reason.

io.agentscope.core.model中Msg的使用

类型变量类型为Msg的io.agentscope.core.model中的方法参数

修饰符和类型	方法	说明
ChatResponse	OllamaChatModel.chat(List<Msg> messages, GenerateOptions options)	Chat with the model using default options.
ChatResponse	OllamaChatModel.chat(List<Msg> messages, OllamaOptions options)	Chat with the model using Ollama-specific options.
protected reactor.core.publisher.Flux<ChatResponse>	AnthropicChatModel.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Stream chat completion responses from Anthropic's API.
protected abstract reactor.core.publisher.Flux<ChatResponse>	ChatModelBase.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Internal implementation for streaming chat completion responses.
protected reactor.core.publisher.Flux<ChatResponse>	DashScopeChatModel.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Stream chat completion responses from DashScope's API.
protected reactor.core.publisher.Flux<ChatResponse>	GeminiChatModel.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Stream chat completion responses from Gemini's API.
protected reactor.core.publisher.Flux<ChatResponse>	OllamaChatModel.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)
protected reactor.core.publisher.Flux<ChatResponse>	OpenAIChatModel.doStream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)
protected reactor.core.publisher.Flux<ChatResponse>	OpenAIChatModel.doStream0(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)
final reactor.core.publisher.Flux<ChatResponse>	ChatModelBase.stream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Stream chat completion responses.
reactor.core.publisher.Flux<ChatResponse>	Model.stream(List<Msg> messages, List<ToolSchema> tools, GenerateOptions options)	Stream chat completion responses.

io.agentscope.core.pipeline中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.pipeline中的方法

修饰符和类型	方法	说明
static Pipeline<Msg>	Pipelines.compose(SequentialPipeline first, SequentialPipeline second)	Compose two sequential pipelines into a single pipeline.
reactor.core.publisher.Mono<List<Msg>>	FanoutPipeline.execute(Msg input)
reactor.core.publisher.Mono<List<Msg>>	FanoutPipeline.execute(Msg input, Class<?> structuredOutputClass)
reactor.core.publisher.Mono<Msg>	SequentialPipeline.execute(Msg input)
reactor.core.publisher.Mono<Msg>	SequentialPipeline.execute(Msg input, Class<?> structuredOutputClass)
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents)	Execute agents in a fanout pipeline with concurrent execution and no input.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents, Msg input)	Execute agents in a fanout pipeline with concurrent execution.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with concurrent execution and structured output.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with concurrent execution, structured output, and no input.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents)	Execute agents in a fanout pipeline with sequential execution and no input.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents, Msg input)	Execute agents in a fanout pipeline with sequential execution.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with sequential execution and structured output.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with sequential execution, structured output, and no input.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents)	Execute agents in a sequential pipeline with no initial input.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents, Msg input)	Execute agents in a sequential pipeline.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a sequential pipeline with structured output.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents, Class<?> structuredOutputClass)	Execute agents in a sequential pipeline with structured output and no initial input.

参数类型为Msg的io.agentscope.core.pipeline中的方法

修饰符和类型	方法	说明
MsgHub.Builder	MsgHub.Builder.announcement(Msg... announcement)	Set announcement messages to broadcast when entering the hub.
reactor.core.publisher.Mono<Void>	MsgHub.broadcast(Msg msg)	Broadcast a message to all participants.
reactor.core.publisher.Mono<List<Msg>>	FanoutPipeline.execute(Msg input)
reactor.core.publisher.Mono<List<Msg>>	FanoutPipeline.execute(Msg input, Class<?> structuredOutputClass)
reactor.core.publisher.Mono<T>	Pipeline.execute(Msg input)	Execute the pipeline with the given input message.
reactor.core.publisher.Mono<T>	Pipeline.execute(Msg input, Class<?> structuredOutputClass)	Execute the pipeline with the given input message and structured output.
reactor.core.publisher.Mono<Msg>	SequentialPipeline.execute(Msg input)
reactor.core.publisher.Mono<Msg>	SequentialPipeline.execute(Msg input, Class<?> structuredOutputClass)
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents, Msg input)	Execute agents in a fanout pipeline with concurrent execution.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanout(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with concurrent execution and structured output.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents, Msg input)	Execute agents in a fanout pipeline with sequential execution.
static reactor.core.publisher.Mono<List<Msg>>	Pipelines.fanoutSequential(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a fanout pipeline with sequential execution and structured output.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents, Msg input)	Execute agents in a sequential pipeline.
static reactor.core.publisher.Mono<Msg>	Pipelines.sequential(List<AgentBase> agents, Msg input, Class<?> structuredOutputClass)	Execute agents in a sequential pipeline with structured output.
reactor.core.publisher.Flux<Event>	FanoutPipeline.stream(Msg input)	Stream execution events from all agents with default options.
reactor.core.publisher.Flux<Event>	FanoutPipeline.stream(Msg input, StreamOptions options)	Stream execution events from all agents with specified options.
reactor.core.publisher.Flux<Event>	FanoutPipeline.stream(Msg input, StreamOptions options, Class<?> structuredOutputClass)	Stream execution events from all agents with structured output support.

类型变量类型为Msg的io.agentscope.core.pipeline中的方法参数

修饰符和类型	方法	说明
MsgHub.Builder	MsgHub.Builder.announcement(List<Msg> announcement)	Set announcement messages to broadcast when entering the hub.
reactor.core.publisher.Mono<Void>	MsgHub.broadcast(List<Msg> msgs)	Broadcast multiple messages to all participants.

io.agentscope.core.plan中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.plan中的方法

修饰符和类型	方法	说明
reactor.core.publisher.Mono<Msg>	PlanNotebook.getCurrentHint()	Gets the current hint message based on plan state.

io.agentscope.core.rag.model中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.rag.model中的方法

修饰符和类型	方法	说明
List<Msg>	RetrieveConfig.getConversationHistory()	Gets the conversation history for context-aware retrieval.

类型变量类型为Msg的io.agentscope.core.rag.model中的方法参数

修饰符和类型	方法	说明
RetrieveConfig.Builder	RetrieveConfig.Builder.conversationHistory(List<Msg> conversationHistory)	Sets the conversation history for context-aware retrieval.

io.agentscope.core.tool中Msg的使用

返回Msg的io.agentscope.core.tool中的方法

修饰符和类型	方法	说明
static Msg	ToolResultMessageBuilder.buildToolResultMsg(ToolResultBlock result, ToolUseBlock originalCall, String agentName)	Build a tool result message from a ToolResultBlock and the original tool call.

参数类型为Msg的io.agentscope.core.tool中的方法

修饰符和类型	方法	说明
static void	ToolValidator.validateToolResultMatch(Msg assistantMsg, List<Msg> inputMsgs)	Validate that ToolResult messages match pending ToolUse blocks.

类型变量类型为Msg的io.agentscope.core.tool中的方法参数

修饰符和类型	方法	说明
static void	ToolValidator.validateToolResultMatch(Msg assistantMsg, List<Msg> inputMsgs)	Validate that ToolResult messages match pending ToolUse blocks.

io.agentscope.core.tracing中Msg的使用

返回变量类型为Msg的类型的io.agentscope.core.tracing中的方法

修饰符和类型	方法	说明
default reactor.core.publisher.Mono<Msg>	Tracer.callAgent(AgentBase instance, List<Msg> inputMessages, Supplier<reactor.core.publisher.Mono<Msg>> agentCall)

类型变量类型为Msg的io.agentscope.core.tracing中的方法参数

修饰符和类型	方法	说明
default reactor.core.publisher.Mono<Msg>	Tracer.callAgent(AgentBase instance, List<Msg> inputMessages, Supplier<reactor.core.publisher.Mono<Msg>> agentCall)
default reactor.core.publisher.Mono<Msg>	Tracer.callAgent(AgentBase instance, List<Msg> inputMessages, Supplier<reactor.core.publisher.Mono<Msg>> agentCall)
default <TReq, TResp, TParams> List<TReq>	Tracer.callFormat(AbstractBaseFormatter<TReq,TResp,TParams> formatter, List<Msg> msgs, Supplier<List<TReq>> formatCall)
default reactor.core.publisher.Flux<ChatResponse>	Tracer.callModel(ChatModelBase instance, List<Msg> inputMessages, List<ToolSchema> toolSchemas, GenerateOptions options, Supplier<reactor.core.publisher.Flux<ChatResponse>> modelCall)

io.agentscope.core.util中Msg的使用

类型变量类型为Msg的io.agentscope.core.util中的方法参数

修饰符和类型	方法	说明
static List<ToolUseBlock>	MessageUtils.extractRecentToolCalls(List<Msg> messages, String agentName)	Extract tool calls from the most recent assistant message in the message list.