Module: core/completion

This module provides the core completion components for AI.JSX.

Interfaces

• FunctionDefinition
• ModelProps
• PlainFunctionParameter

Type Aliases

FunctionParameters

Ƭ FunctionParameters: Record<string, PlainFunctionParameter> | z.ZodObject<any>

Represents parameters to a FunctionDefinition.

This type allows two ways for specifying parameters:

• For simple use cases, a record of parameter names to PlainFunctionParameter objects.
• For more complex use cases, a z.ZodObject schema object (zod is a standard runtime type definition & checking library).

Note

If using a Zod schema, the top-level schema must be an object as per OpenAI specifications: https://platform.openai.com/docs/api-reference/chat/create#chat/create-parameters

For example, to describe a list of strings, the following is not accepted: const schema: z.Schema = z.array(z.string())

Instead, you can wrap it in an object like so: const schema: z.ZodObject = z.object({ arr: z.array(z.string()) })

Defined in

packages/ai-jsx/src/core/completion.tsx:98

---

ModelComponent

Ƭ ModelComponent<T>: Component<T>

A Component that invokes a Large Language Model.

Type parameters

Name	Type
T	extends ModelPropsWithChildren

Defined in

packages/ai-jsx/src/core/completion.tsx:38

---

ModelPropsWithChildren

Ƭ ModelPropsWithChildren: ModelProps & { children: Node }

Represents a ModelProps with child @{link Node}s.

Defined in

packages/ai-jsx/src/core/completion.tsx:31

Functions

AssistantMessage

▸ AssistantMessage(«destructured»): Node

Provide an Assistant Message to the LLM, for use within a ChatCompletion.

The assistant message tells the model what it has previously said. See https://platform.openai.com/docs/guides/gpt/chat-completions-api for more detail.

Example

   <ChatCompletion>
     <UserMessage>I'd like to cancel my account.</UserMessage>
     <AssistantMessage>Sorry to hear that. Can you tell me why?</AssistantMessage>
     <UserMessage>It's too expensive.</UserMessage>
   </ChatCompletion>

==> "Ok, thanks for that feedback. I'll cancel your account."

Parameters

Name	Type
«destructured»	Object
› children	Node

Returns

Node

Defined in

packages/ai-jsx/src/core/completion.tsx:258

---

ChatCompletion

▸ ChatCompletion(«destructured», «destructured»): Element

Perform a Large Language Model call to do chat completion.

Every child of ChatCompletion must something that renders to a SystemMessage, UserMessage, or AssistantMessage.

Example

   function MyUserMessage() {
    return <UserMessage>Hi, I'm a user message.</UserMessage>;
   }

   <ChatCompletion>
     <SystemMessage>You are a nice person.</SystemMessage>
     <MyUserMessage />
   </ChatCompletion>

Parameters

Name	Type
«destructured»	ModelProps & { children: Node } & Record<string, unknown>
«destructured»	RenderContext

Returns

Element

Defined in

packages/ai-jsx/src/core/completion.tsx:367

---

ChatProvider

▸ ChatProvider<T>(«destructured», «destructured»): Element

A ChatProvider is used by ChatCompletion to access an underlying Large Language Model.

Type parameters

Name	Type
T	extends ModelPropsWithChildren

Parameters

Name	Type
«destructured»	{ component?: ModelComponent<T> } & T
«destructured»	RenderContext

Returns

Element

Defined in

packages/ai-jsx/src/core/completion.tsx:191

---

Completion

▸ Completion(«destructured», «destructured»): Element

Perform a Large Language Mokdel call to do a completion.

In general, you should prefer to use ChatCompletion instead of Completion, because ChatCompletion uses better models.

Example

   <Completion>
     Here's a list of three dog names:
   </Completion>

   ==> 'Dottie, Murphy, Lucy'

Parameters

Name	Type
«destructured»	ModelProps & { children: Node } & Record<string, unknown>
«destructured»	RenderContext

Returns

Element

Defined in

packages/ai-jsx/src/core/completion.tsx:338

---

CompletionProvider

▸ CompletionProvider<T>(«destructured», «destructured»): Element

A CompletionProvider is used by ChatCompletion to access an underlying Large Language Model.

Type parameters

Name	Type
T	extends ModelPropsWithChildren

Parameters

Name	Type
«destructured»	{ component?: ModelComponent<T> } & T
«destructured»	RenderContext

Returns

Element

Defined in

packages/ai-jsx/src/core/completion.tsx:168

---

ConversationHistory

▸ ConversationHistory(«destructured»): Element[]

Parameters

Name	Type
«destructured»	Object
› messages	ChatCompletionResponseMessage[]

Returns

Element[]

Defined in

packages/ai-jsx/src/core/completion.tsx:262

---

FunctionCall

▸ FunctionCall(«destructured»): string

Provide a function call to the LLM, for use within a ChatCompletion.

The function call tells the model that a function was previously invoked by the model. See https://platform.openai.com/docs/guides/gpt/chat-completions-api for more detail. When the model returns a function call, @{link ChatCompletion} returns a @{link FunctionCall} component.

Example

   <ChatCompletion>
     <UserMessage>What is 258 * 322?</UserMessage>
     <FunctionCall name="evaluate_math" args={expression: "258 * 322"} />
     <FunctionResponse name="evaluate_math">83076</FunctionResponse>
   </ChatCompletion>

   ==> "That would be 83,076."

Parameters

Name	Type
«destructured»	Object
› args	Record<string, null | string | number | boolean>
› name	string

Returns

string

Defined in

packages/ai-jsx/src/core/completion.tsx:296

---

FunctionResponse

▸ FunctionResponse(«destructured», «destructured»): Promise<string>

Renders to the output of a previous FunctionCall component, for use within a ChatCompletion.

See https://platform.openai.com/docs/guides/gpt/chat-completions-api for more detail.

Example

   <ChatCompletion>
     <UserMessage>What is 258 * 322?</UserMessage>
     <FunctionCall name="evaluate_math" args={expression: "258 * 322"} />
     <FunctionResponse name="evaluate_math">83076</FunctionResponse>
   </ChatCompletion>

   ==> "That would be 83,076."

Parameters

Name	Type
«destructured»	Object
› children	Node
› name	string
«destructured»	ComponentContext

Returns

Promise<string>

Defined in

packages/ai-jsx/src/core/completion.tsx:316

---

SystemMessage

▸ SystemMessage(«destructured»): Node

Provide a System Message to the LLM, for use within a ChatCompletion.

The system message can be used to put the model in character. See https://platform.openai.com/docs/guides/gpt/chat-completions-api for more detail.

Example

   <ChatCompletion>
     <SystemMessage>You are a helpful customer service agent.</SystemMessage>
   </ChatCompletion>

Parameters

Name	Type
«destructured»	Object
› children	Node

Returns

Node

Defined in

packages/ai-jsx/src/core/completion.tsx:220

---

UserMessage

▸ UserMessage(«destructured»): Node

Provide a User Message to the LLM, for use within a ChatCompletion.

The user message tells the model what the user has said. See https://platform.openai.com/docs/guides/gpt/chat-completions-api for more detail.

Example

   <ChatCompletion>
     <UserMessage>I'd like to cancel my account.</UserMessage>
   </ChatCompletion>

   ==> 'Sorry to hear that. Can you tell me why?

Parameters

Name	Type
«destructured»	Object
› children	Node
› name?	string

Returns

Node

Defined in

packages/ai-jsx/src/core/completion.tsx:238

---

getParametersSchema

▸ getParametersSchema(parameters): JsonSchema7Meta & {} | { properties: {} ; required: string[] ; type: string = 'object' }

This function creates a JSON Schema object to describe parameters for a FunctionDefinition.

See FunctionParameters for more information on what parameters are supported.

Parameters

Name	Type
parameters	FunctionParameters

Returns

JsonSchema7Meta & {} | { properties: {} ; required: string[] ; type: string = 'object' }

Defined in

packages/ai-jsx/src/core/completion.tsx:54