Class DatalayerChatParticipant

Chat participant that provides context from Datalayer notebooks and lexical documents

Provides interactive assistance with tool invocation for working with Jupyter notebooks and Lexical documents.

Uses Datalayer tools to interact with notebooks and lexical documents. Always calls getActiveDocument first, then listAvailableBlocks for lexical documents.

Example

@datalayer connect to pyodide, insert a fibonacci cell, and run all cells
@datalayer explain this notebook
@datalayer what cells are in this document?

Index

Constructors

constructor

new DatalayerChatParticipant(
context: ExtensionContext,
): DatalayerChatParticipant
Parameters
- context: ExtensionContext
Returns DatalayerChatParticipant
- Defined in src/chat/datalayerChatParticipant.ts:22

Properties

`Private`context

context: ExtensionContext

`Private`participant

participant: ChatParticipant

Methods

`Private`buildSimpleSystemPrompt

buildSimpleSystemPrompt(): string
Build a simple system prompt that instructs the model to use Datalayer tools

Returns string
- Defined in src/chat/datalayerChatParticipant.ts:291

`Private`convertHistoryToMessages

convertHistoryToMessages(
history: readonly (ChatRequestTurn | ChatResponseTurn)[],
): LanguageModelChatMessage[]
Convert VS Code chat history to language model messages. Extracts user prompts and assistant responses from previous turns.
Parameters
- history: readonly (ChatRequestTurn | ChatResponseTurn)[]
Returns LanguageModelChatMessage[]
- Defined in src/chat/datalayerChatParticipant.ts:259

`Private`handleChatRequest

handleChatRequest(
    request: ChatRequest,
    _context: ChatContext,
    stream: ChatResponseStream,
    token: CancellationToken,
): Promise<void>
Parameters
- request: ChatRequest
- _context: ChatContext
- stream: ChatResponseStream
- token: CancellationToken
Returns Promise<void>
- Defined in src/chat/datalayerChatParticipant.ts:44

`Private`processModelResponse

processModelResponse(
    model: LanguageModelChat,
    messages: LanguageModelChatMessage[],
    tools: LanguageModelChatTool[],
    stream: ChatResponseStream,
    request: ChatRequest,
    token: CancellationToken,
): Promise<void>
Process model response and handle tool calls in a loop
Parameters
- model: LanguageModelChat
- messages: LanguageModelChatMessage[]
- tools: LanguageModelChatTool[]
- stream: ChatResponseStream
- request: ChatRequest
- token: CancellationToken
Returns Promise<void>
- Defined in src/chat/datalayerChatParticipant.ts:174

Class DatalayerChatParticipant

Example

Index

Constructors

Properties

Methods

Constructors

constructor

Parameters

Returns DatalayerChatParticipant

Properties

`Private`context

`Private`participant

Methods

`Private`buildSimpleSystemPrompt

Returns string

`Private`convertHistoryToMessages

Parameters

Returns LanguageModelChatMessage[]

`Private`handleChatRequest

Parameters

Returns Promise<void>

`Private`processModelResponse

Parameters

Returns Promise<void>

register

Returns Disposable

Settings

On This Page

Class DatalayerChatParticipant

Example

Index

Constructors

Properties

Methods

Constructors

constructor

Parameters

Returns DatalayerChatParticipant

Properties

Privatecontext

Privateparticipant

Methods

PrivatebuildSimpleSystemPrompt

Returns string

PrivateconvertHistoryToMessages

Parameters

Returns LanguageModelChatMessage[]

PrivatehandleChatRequest

Parameters

Returns Promise<void>

PrivateprocessModelResponse

Parameters

Returns Promise<void>

register

Returns Disposable

Settings

On This Page

`Private`context

`Private`participant

`Private`buildSimpleSystemPrompt

`Private`convertHistoryToMessages

`Private`handleChatRequest

`Private`processModelResponse