Skip to content

Getting Started

This guide gets you from zero to your first Ergo query in under 2 minutes.

Installation

pip install ergo-agent

For LLM framework integration, install the extras you need:

pip install ergo-agent[openai]

pip install ergo-agent[anthropic]

pip install ergo-agent[langchain]

pip install ergo-agent[all]

Mode 1: Read-Only (no wallet)

Perfect for price bots, portfolio trackers, and exploration.

from ergo_agent import ErgoNode, Wallet
from ergo_agent.tools import ErgoToolkit

node = ErgoNode()  # uses public API — no node required
wallet = Wallet.read_only("9f4QF8jQSBiHrgqrCDuS3L62MY6MaBFW5UeqNqfEi1mCfmPFxVo")
toolkit = ErgoToolkit(node=node, wallet=wallet)

# Check balance
balance = toolkit.get_wallet_balance()
print(balance)

# Get live ERG/USD price
price = toolkit.get_erg_price()
print(price)

No API key needed

The SDK uses the public api.ergoplatform.com by default. Rate limits are generous for development use.

Mode 2: Node Wallet (transactions enabled)

For sending ERG, swapping tokens, and signing transactions.

from ergo_agent import ErgoNode, Wallet
from ergo_agent.tools import ErgoToolkit, SafetyConfig

# Connect to your own Ergo node
node = ErgoNode(node_url="http://localhost:9053", api_key="your-api-key")
wallet = Wallet.from_node_wallet("9f...")

toolkit = ErgoToolkit(
    node=node,
    wallet=wallet,
    safety=SafetyConfig(
        max_erg_per_tx=5.0,       # hard cap per transaction
        max_erg_per_day=50.0,     # rolling 24h limit
        rate_limit_per_hour=20,   # max 20 actions/hour
    ),
)

# Send ERG (passes through safety checks)
result = toolkit.send_erg(to="9f...", amount_erg=1.5)

Node required

Transaction signing requires connecting to an Ergo node with the wallet API unlocked. See the Ergo node setup guide for instructions.

Mode 3: LLM Agent Integration

Connect any LLM to the Ergo blockchain with tool/function calling.

from ergo_agent import ErgoNode, Wallet
from ergo_agent.tools import ErgoToolkit

node = ErgoNode()
wallet = Wallet.read_only("9f...")
toolkit = ErgoToolkit(node=node, wallet=wallet)

# Generate OpenAI function-calling tool definitions
tools = toolkit.to_openai_tools()
# Pass `tools` to your OpenAI chat completion call

tools = toolkit.to_anthropic_tools()
# Pass `tools` to your Anthropic messages call

lc_tools = toolkit.to_langchain_tools()
# Use with LangChain's AgentExecutor or create_tool_calling_agent

When the LLM calls a tool, execute it:

result = toolkit.execute_tool("get_erg_price", {})
# Returns JSON string: {"erg_usd": 0.31, "source": "oracle_pool_v2"}

What's Next?