Crypto.com AI Agent SDK
  • Getting Started
    • Introduction
    • Quick Start Guide: Simulation Entry Point
  • Core Concepts Overview
    • Plugins
    • Tools
    • Blockchain Functions
      • Wallet Management
      • Token Interaction
      • Transaction Queries
      • Block Information
      • Contract Operations
      • CronosID Operations
      • Defi Operations
      • Crypto.com Exchange
    • Built-in Telegram Feature
    • Advanced Usage: Custom Instructions
    • Dynamic AI Model Manager
  • Crypto.com On-Chain Developer Platform
    • Developer Platform API Methods
  • On-Chain Functions and Examples
    • AI Agent Chatbot
    • "Magic Link" Signer
  • Resources for Developers
Powered by GitBook
On this page
  • Storage Plugin
  • LangFuse Plugin
  • Personality Plugin

Was this helpful?

  1. Core Concepts Overview

Plugins

Plugins allow you to extend the functionality of the agent. Below are the available plugins and examples of their usage.

Storage Plugin

The SQLitePlugin provides a mechanism for state persistence using SQLite.

A critical persistence component for AI Agents that leverages SQLite's robust database engine to maintain state and memory across interactions. This plugin enables agents to store and recall important information, conversation history, and learned patterns using SQLite's self-contained, serverless architecture, ensuring reliable data management without external database dependencies.

Disclaimer: Cryptocom AI Agent SDK does not provide built-in encryption for data stored via SQLitePlugin. Developers are fully responsible for securing any data written to disk using this plugin. Crypto.com is not liable for any data leakage or compromise resulting from insecure use of this storage method.

Best Practices

For production use, developers are responsible for ensuring secure data handling. Key guideline to follow:

  • Do not store private keys, secrets, or personal user data in the agent state unless encryption is implemented.

  • Consider using encrypted database extensions (e.g., SQLCipher) or manually encrypting sensitive fields before storage.

  • Store the database .db file in secure, access-controlled directories to prevent unauthorized access.

  • If your application is running on user devices or servers, utilize OS-level encryption and enforce strict permissions to safeguard data.

  • Transmit data over secure connections to protect data in transit (e.g., HTTPS).

Example Usage

from crypto_com_agent_client import Agent, SQLitePlugin

agent = Agent.init(
    llm_config={
        "provider": "OpenAI",
        "model": "gpt-4",
        "provider-api-key": "sk-proj-example-key",
        "temperature": "float-controlling-output-randomness",
    },
    blockchain_config={
         "chainId": "chain-id",
         "explorer-api-key": "your-api-key",
         "private-key": "your-private-key",
         "sso-wallet-url": "your-sso-wallet-url",
    },
    plugins={                
        "storage": SQLitePlugin(db_path="agent_state.db"),
    },
)

LangFuse Plugin

The LangFusePlugin integrates telemetry for monitoring interactions.

An observability and monitoring solution specifically designed for AI Agents, providing detailed telemetry of agent-user interactions and performance metrics. This open-source plugin enables development teams to track, debug, and optimize their agent's behavior through comprehensive analytics and collaborative tools, making it easier to identify areas for improvement and ensure optimal agent performance.

Provide the required keys as a dictionary.

Example Usage

from crypto_com_agent_client import Agent

agent = Agent.init(
    llm_config={
        "provider": "OpenAI",
        "model": "gpt-4",
        "provider-api-key": "sk-proj-example-key",
        "temperature": "float-controlling-output-randomness",
    },
    blockchain_config={
         "chainId": "chain-id",
         "explorer-api-key": "your-api-key",
         "private-key": "your-private-key",
         "sso-wallet-url": "your-sso-wallet-url",
    },
    plugins={
        "langfuse": {
            "public-key": "user-public-key",
            "secret-key": "user-secret-key",
            "host": "https://langfuse.example.com",
        }
    },
)

Personality Plugin

Customize the agent’s personality and instructions. A configuration module that enables fine-tuned control over an agent's communication style and behavior patterns.

Supported Keys:

  • Tone: e.g., "friendly", "professional", "humorous"

  • Language: e.g., "English", "German", "Spanish"

  • Verbosity: e.g., "low", "medium", "high"

These keys allow developers to easily control the agent's communication style to suit specific use cases or brand requirements, ensuring consistent and appropriate interactions across all scenarios.

Example Usage

from crypto_com_agent_client import Agent

agent = Agent.init(
    llm_config={
        "provider": "OpenAI",
        "model": "gpt-4",
        "provider-api-key": "sk-proj-example-key",
        "temperature": "float-controlling-output-randomness",
    },
    blockchain_config={
         "chainId": "chain-id",
         "explorer-api-key": "your-api-key",
         "private-key": "your-private-key",
         "sso-wallet-url": "your-sso-wallet-url",
    },
    plugins={
        "personality": {
            "tone": "friendly",
            "language": "English",
            "verbosity": "high",
        },
    },
)

For more advanced behavior control, developers can combine personality configuration with freeform instructions. This enables deeper modification of the AI Agent’s behavior for complex and nuanced personality traits that enhance task performance and user interactions.


PreviousCore Concepts OverviewNextTools

Last updated 2 days ago

Was this helpful?

Learn more in the .

Advanced Usage