agent-provider-google / Exports / GoogleProvider

Class: GoogleProvider

Google Gemini provider implementation for Robota

IMPORTANT PROVIDER-SPECIFIC RULES:

1. This provider MUST extend BaseAIProvider from @robota-sdk/agent-core
2. Content handling for Google Gemini API:
   • Function calls can have content (text) along with function calls
   • Content can be empty string or actual text, NOT null
3. Use override keyword for all methods inherited from BaseAIProvider
4. Provider-specific API behavior should be documented here

Hierarchy

• AbstractAIProvider

  ↳ GoogleProvider

Implements

• IImageGenerationProvider

Table of contents

Constructors

• constructor

Properties

• name
• version

Methods

• configure
• generateResponse
• generateStreamingResponse
• chat
• chatStream
• generateImage
• editImage
• composeImage
• supportsTools
• validateConfig
• dispose

Constructors

constructor

• new GoogleProvider(options): GoogleProvider

Parameters

Name	Type
options	IGoogleProviderOptions

Returns

GoogleProvider

Overrides

AbstractAIProvider.constructor

Defined in

agent-provider-google/src/provider.ts:48

Properties

name

• Readonly name: "google"

Overrides

AbstractAIProvider.name

Defined in

agent-provider-google/src/provider.ts:42

---

version

• Readonly version: "1.0.0"

Overrides

AbstractAIProvider.version

Defined in

agent-provider-google/src/provider.ts:43

Methods

configure

▸ configure(config): Promise<void>

Configure the provider with type-safe configuration

Parameters

Name	Type
config	IProviderConfig

Returns

Promise<void>

Inherited from

AbstractAIProvider.configure

Defined in

agent-core/dist/node/index.d.ts:2464

---

generateResponse

▸ generateResponse(payload): Promise<IRawProviderResponse>

Provider-agnostic raw response API.

This is the canonical "raw payload" entrypoint required by the AIProvider contract. The default implementation delegates to chat() and adapts the result into a RawProviderResponse shape.

Parameters

Name	Type
payload	IProviderRequest

Returns

Promise<IRawProviderResponse>

Inherited from

AbstractAIProvider.generateResponse

Defined in

agent-core/dist/node/index.d.ts:2487

---

generateStreamingResponse

▸ generateStreamingResponse(payload): AsyncIterable<IRawProviderResponse, any, any>

Provider-agnostic raw streaming API.

If a provider does not implement chatStream, it does not support streaming.

Parameters

Name	Type
payload	IProviderRequest

Returns

AsyncIterable<IRawProviderResponse, any, any>

Inherited from

AbstractAIProvider.generateStreamingResponse

Defined in

agent-core/dist/node/index.d.ts:2493

---

chat

▸ chat(messages, options?): Promise<TUniversalMessage>

Generate response using TUniversalMessage

Parameters

Name	Type
messages	TUniversalMessage[]
options?	IChatOptions

Returns

Promise<TUniversalMessage>

Overrides

AbstractAIProvider.chat

Defined in

agent-provider-google/src/provider.ts:62

---

chatStream

▸ chatStream(messages, options?): AsyncIterable<TUniversalMessage, any, any>

Generate streaming response using TUniversalMessage

Parameters

Name	Type
messages	TUniversalMessage[]
options?	IChatOptions

Returns

AsyncIterable<TUniversalMessage, any, any>

Overrides

AbstractAIProvider.chatStream

Defined in

agent-provider-google/src/provider.ts:89

---

generateImage

▸ generateImage(request): Promise<TProviderMediaResult<IImageGenerationResult>>

Generate an image from a text prompt using the Gemini API.

Parameters

Name	Type
request	IImageGenerationRequest

Returns

Promise<TProviderMediaResult<IImageGenerationResult>>

Implementation of

IImageGenerationProvider.generateImage

Defined in

agent-provider-google/src/provider.ts:116

---

editImage

▸ editImage(request): Promise<TProviderMediaResult<IImageGenerationResult>>

Edit an existing image based on a text prompt using the Gemini API.

Parameters

Name	Type
request	IImageEditRequest

Returns

Promise<TProviderMediaResult<IImageGenerationResult>>

Implementation of

IImageGenerationProvider.editImage

Defined in

agent-provider-google/src/provider.ts:148

---

composeImage

▸ composeImage(request): Promise<TProviderMediaResult<IImageGenerationResult>>

Compose multiple images together based on a text prompt using the Gemini API.

Parameters

Name	Type
request	IImageComposeRequest

Returns

Promise<TProviderMediaResult<IImageGenerationResult>>

Implementation of

IImageGenerationProvider.composeImage

Defined in

agent-provider-google/src/provider.ts:185

---

supportsTools

▸ supportsTools(): boolean

Returns

boolean

Overrides

AbstractAIProvider.supportsTools

Defined in

agent-provider-google/src/provider.ts:235

---

validateConfig

▸ validateConfig(): boolean

Returns

boolean

Overrides

AbstractAIProvider.validateConfig

Defined in

agent-provider-google/src/provider.ts:239

---

dispose

▸ dispose(): Promise<void>

Returns

Promise<void>

Overrides

AbstractAIProvider.dispose

Defined in

agent-provider-google/src/provider.ts:243