prompt-cache-model/ai.koog.prompt.cache.model/PromptCache

PromptCache

Interface for caching prompt execution results. Implementations should provide a way to store and retrieve prompt execution results.

Inheritors

InMemoryPromptCache

Types

Factory

interface Factory

Represents a factory interface for creating instances of PromptCache. Factories implementing this interface are designed to construct specific types of prompt caches based on a provided configuration string.

Request

@Serializable

class Request

Represents a request to be cached, consisting of a prompt and optional tools. This class is used by PromptCache implementations to store and retrieve cached responses.

Functions

get

abstract suspend fun get(request: PromptCache.Request): List<Message.Response>?

Get a cached response for a request, or null if not cached.

get

suspend fun PromptCache.get(prompt: Prompt, tools: List<ToolDescriptor>, clock: Clock = Clock.System): List<Message.Response>?

Get a cached response for a prompt with tools, or null if not cached.

put

abstract suspend fun put(request: PromptCache.Request, response: List<Message.Response>)

Put a response in the cache for a request.

put

suspend fun PromptCache.put(prompt: Prompt, tools: List<ToolDescriptor>, response: List<Message.Response>)

Put a response in the cache for a prompt with tools.