Solana Agent Kit provides comprehensive integration with Allora’s API for accessing price inferences and topics. The integration supports both mainnet and testnet environments with configurable API settings.

​
Key Features

  • Price inference for multiple tokens and timeframes
  • Topic management and retrieval
  • Inference data by topic ID
  • LangChain tool integration
  • Configurable network environment

​
Configuration

The Allora integration uses the following configuration parameters: 
const config: AlloraAPIClientConfig = {
  apiKey: string,         // Your Allora API key
  chainSlug: ChainSlug,  // MAINNET or TESTNET
  baseAPIUrl: string     // Optional custom API URL
};

​
Basic Usage

​
Getting All Topics

const topics = await agent.methods.getAllTopics();

​
Getting Inference by Topic ID

const inference = await agent.methods.getInferenceByTopicId(topicId);

​
Getting Price Inference

const priceInference = await agent.methods.getPriceInference(
  tokenSymbol,  // e.g., "BTC"
  timeframe     // e.g., "5m"
);

​
LangChain Integration

Solana Agent Kit provides LangChain tools for automated Allora data retrieval:

​
Get All Topics Tool

import { SolanaAlloraGetAllTopics } from 'solana-agent-kit';

const getAllTopicsTool = new SolanaAlloraGetAllTopics(agent);

// Tool returns JSON response:
{
  status: "success",
  message: "Topics fetched successfully",
  topics: AlloraTopic[]
}

​
Get Inference by Topic ID Tool

import { SolanaAlloraGetInferenceByTopicId } from 'solana-agent-kit';

const getInferenceByTopicIdTool = new SolanaAlloraGetInferenceByTopicId(agent);

// Tool input: topic ID as string
const input = "42";

// Tool returns JSON response:
{
  status: "success",
  message: "Inference fetched successfully",
  topicId: number,
  inference: AlloraInference
}

​
Get Price Inference Tool

import { SolanaAlloraGetPriceInference } from 'solana-agent-kit';

const getPriceInferenceTool = new SolanaAlloraGetPriceInference(agent);

// Tool input format (JSON string):
const input = JSON.stringify({
  tokenSymbol: "BTC",
  timeframe: "5m"
});

// Tool returns JSON response:
{
  status: "success",
  message: "Price inference fetched successfully",
  tokenSymbol: string,
  timeframe: string,
  priceInference: string
}

​
Example Prompts

For LangChain AI tools, here are example prompts:

​
Getting Topics

"Get all available Allora topics"
"List the current topics from Allora"

​
Getting Inferences

"Get the inference for topic ID 42"
"What's the current inference for BTC with 5-minute timeframe"

​
Important Notes

  1. API Key Configuration
    • Default API key provided but recommended to use your own
    • Can be configured via ALLORA_API_KEY environment variable
  2. Network Environment
    • Supports both mainnet and testnet
    • Configured via ALLORA_NETWORK environment variable
    • Defaults to testnet if not specified
  3. API Response Handling
    • All methods return structured responses
    • Error handling included for network and API issues
    • Status and message fields in all responses

​
Error Handling

try {
  await agent.methods.getAllTopics();
} catch (error) {
  if (error.message.includes("API key invalid")) {
    // Handle authentication error
  } else if (error.message.includes("network error")) {
    // Handle connection issues
  }
}

​
Technical Details

​
Available Timeframes

enum PriceInferenceTimeframe {
  FIVE_MIN = "5m",
  FIFTEEN_MIN = "15m",
  THIRTY_MIN = "30m",
  ONE_HOUR = "1h",
  FOUR_HOUR = "4h",
  ONE_DAY = "1d"
}

​
Supported Token Types

enum PriceInferenceToken {
  BTC = "BTC",
  ETH = "ETH",
  SOL = "SOL"
  // Additional tokens based on API support
}

​
Default Values

const DEFAULT_CONFIG = {
  ALLORA_API_KEY: "UP-d33e797de5134909854be2b7",
  ALLORA_NETWORK: "testnet"
};