Fuzzing API

fuzz

@boa.fuzz(contract_function)

Derive Hypothesis strategies from a deployed Vyper function's canonical ABI argument types and apply them to the decorated Python test.

import boa

contract = boa.loads(
    """
@external
def identity(item: uint256) -> uint256:
    return item
"""
)


@boa.fuzz(contract.identity)
def check_identity(item):
    assert contract.identity(item) == item


check_identity()

The result is a Hypothesis test and must be called (or collected by pytest). A strategy is generated for every Vyper argument, including arguments with defaults. Use boa.test.strategy with hypothesis.given when you need custom ranges or sizes.

strategy

boa.test.strategy(type_string, **kwargs)

Return a Hypothesis strategy for a canonical ABI type string.

from hypothesis import given

from boa.test import strategy


@given(
    account=strategy("address"),
    amount=strategy("uint256", min_value=1, max_value=10**24),
    route=strategy("address[]", min_length=2, max_length=5),
)
def test_inputs(account, amount, route):
    assert amount > 0
    assert 2 <= len(route) <= 5

Supported values include integers, address, bool, fixed and dynamic bytes, string, arrays, tuples, and Vyper decimal. Type-specific keyword arguments are passed to the underlying strategy builder. Nested dynamic arrays accept a list for min_length or max_length, one value per dynamic dimension. Dynamic bytes defaults to min_size=1. Currently, passing min_size=0 still uses 1; combine the strategy with hypothesis.strategies.just(b"") to include empty bytes.

The decimal strategy generates Python Decimal values. Directly passing them to decimal contract arguments is currently unsupported by Titanoboa's ABI encoder; see the decimal limitations.

See the fuzzing strategies guide for per-type examples, @boa.fuzz, composite and stateful testing, pytest isolation, and common patterns.