sardis-adk 0.1.0

Google Agent Development Kit (ADK) integration for Sardis

sardis-adk

Google Agent Development Kit (ADK) integration for Sardis -- the Payment OS for the Agent Economy.

Give your Google ADK agents the ability to make real payments, check balances, enforce spending policies, and review transaction history through Sardis.

Installation

pip install sardis-adk

Quick Start

Option 1: Pre-configured Agent

The fastest way to get a payment-capable agent:

from sardis_adk import create_sardis_agent

agent = create_sardis_agent(
    api_key="sk_test_...",
    wallet_id="wallet_abc123",
    model="gemini-2.0-flash",
)

# Use with ADK runner
from google.adk.runners import InMemoryRunner

runner = InMemoryRunner(agent=agent)

Option 2: Toolkit (Bring Your Own Agent)

Add Sardis tools to any existing ADK agent:

from google.adk import Agent
from sardis_adk import SardisToolkit

toolkit = SardisToolkit(api_key="sk_test_...", wallet_id="wallet_abc123")

agent = Agent(
    name="my_agent",
    model="gemini-2.0-flash",
    tools=toolkit.get_tools(),
    instruction="You are a helpful assistant that can make payments.",
)

Option 3: Individual Tools

Use specific tools without the toolkit:

from sardis_adk.tools import sardis_pay, sardis_check_balance, configure
from sardis import SardisClient

client = SardisClient(api_key="sk_test_...")
wallet = client.wallets.create(name="my-agent", chain="base", policy="Max $100/day")

configure(client, wallet.id)

# Now use tools directly
result = sardis_pay(to="openai.com", amount="25.00", purpose="API credits")
print(result["success"])  # True

Available Tools

Tool	Description
sardis_pay	Execute a payment with automatic policy enforcement
sardis_check_balance	Check wallet balance and spending limits
sardis_check_policy	Validate whether a payment would be allowed (dry-run)
sardis_set_policy	Set or update spending policy using natural language
sardis_list_transactions	View recent transaction history

Tool Details

sardis_pay

sardis_pay(
    to="openai.com",        # Recipient address or merchant
    amount="25.00",          # Payment amount
    token="USDC",            # Token type (USDC, USDT, PYUSD, EURC)
    purpose="API credits",   # Reason for payment
)
# Returns: {"success": True, "tx_id": "tx_...", "balance_after": 975.0, ...}

sardis_check_balance

sardis_check_balance(token="USDC", chain="base")
# Returns: {"balance": 1000.0, "remaining": 900.0, "limit_per_tx": 100.0, ...}

sardis_check_policy

sardis_check_policy(to="openai.com", amount="25.00", token="USDC")
# Returns: {"allowed": True, "reason": "All policy checks passed", ...}

sardis_set_policy

sardis_set_policy(policy_text="Max $50 per transaction, daily limit $500")
# Returns: {"success": True, "limit_per_tx": 50.0, "limit_total": 500.0, ...}

sardis_list_transactions

sardis_list_transactions(limit=10)
# Returns: {"count": 3, "transactions": [{"tx_id": "tx_...", ...}, ...]}

Simulation Mode

By default, Sardis runs in simulation mode -- no real money moves and no API key is required. This is ideal for development and testing:

from sardis_adk import SardisToolkit

# Simulation mode (no real transactions)
toolkit = SardisToolkit(api_key="sk_test_demo", wallet_id="wallet_abc")
tools = toolkit.get_tools()

When you are ready for production, use a real API key and Sardis will execute on-chain transactions through non-custodial MPC wallets.

Combining with Other Tools

Sardis tools work alongside any other ADK tools:

from google.adk import Agent
from google.adk.tools import google_search
from sardis_adk import SardisToolkit

toolkit = SardisToolkit(api_key="sk_test_...", wallet_id="wallet_abc")

agent = Agent(
    name="research_and_pay",
    model="gemini-2.0-flash",
    tools=[google_search, *toolkit.get_tools()],
    instruction="Research products and make purchases when instructed.",
)

Requirements

• Python 3.11+
• sardis >= 0.3
• google-adk >= 0.3

Links

• Sardis Documentation
• Google ADK Documentation
• GitHub Repository