Unchained Solana Agent Framework

Build powerful AI-powered agent systems for Solana blockchain operations, trading, and analytics

Solana Integration

Seamlessly interact with Solana blockchain using AI-powered agents and smart contracts.

Advanced Orchestration

Coordinate multiple AI agents for complex on-chain operations and data analysis.

Enterprise Security

Production-grade security with built-in wallet management and transaction signing.

Key Features

Solana Integration

Native support for Solana blockchain operations and smart contract interactions

AI-Powered Agents

Intelligent agents for automated trading, monitoring, and analysis

Multi-Agent Systems

Coordinate multiple agents for complex blockchain operations

Advanced Orchestration

Built-in workflow patterns for common Solana operations

Secure by Design

Enterprise-grade security with wallet integration and key management

High Performance

Optimized for high-frequency trading and real-time monitoring

Multi-Agent Structures

Different patterns for organizing multiple agents

Sequential Workflow

from agentify import Agent, SequentialWorkflow
from agentify.models import Anthropic

llm = Anthropic()

agent1 = Agent(
    agent_name="Blog generator",
    system_prompt="Generate a blog post like stephen king",
    llm=llm,
    max_loops=1,
    dashboard=False,
    tools=[],
)

agent2 = Agent(
    agent_name="Summarizer",
    system_prompt="Summarize the blog post",
    llm=llm,
    max_loops=1,
    dashboard=False,
    tools=[],
)

workflow = SequentialWorkflow(
    agents=[agent1, agent2],
    workflow_name="Blog Generation Pipeline"
)

Sequential Workflow

Chain multiple agents together in a sequence, where each agent performs a specific task and passes results to the next agent.

Key Features:

  • Linear processing pipeline
  • Clear data flow between agents
  • Simple to understand and implement
  • Predictable execution order

Multi-Agent Architecture

TaskInput
Layer 1: Reference Agents
Agent 1
Agent 2
Agent N
Agent 1 Response
Agent 2 Response
Agent N Response
Layer 2: Aggregator Agent
Aggregate All Responses
Final Output