Testing and Code Coverage for CosmWasm Contracts
This skill helps you write comprehensive tests and achieve high code coverage for Axone protocol smart contracts.
When to use this skill
Use this skill when you need to:
- •Write unit tests for contract logic
- •Create integration tests with Abstract SDK
- •Set up test environments with cw-orch
- •Analyze and improve code coverage
- •Test error conditions and edge cases
Test Environment Setup
Integration Test Structure
rust
// tests/integration.rs
use my_contract::{
contract::interface::MyContractInterface,
msg::{
MyContractExecuteMsgFns, MyContractInstantiateMsg,
MyContractQueryMsgFns, ConfigResponse,
},
MyContractError, AXONE_NAMESPACE,
};
use abstract_app::objects::namespace::Namespace;
use abstract_client::{AbstractClient, Application, Environment};
use cosmwasm_std::coins;
use cw_orch::{anyhow, prelude::*};
struct TestEnv<Env: CwEnv> {
abs: AbstractClient<Env>,
app: Application<Env, MyContractInterface<Env>>,
}
impl TestEnv<MockBech32> {
fn setup() -> anyhow::Result<TestEnv<MockBech32>> {
let mock = MockBech32::new("mock");
let sender = mock.sender_addr();
let namespace = Namespace::new(AXONE_NAMESPACE)?;
// Build Abstract client
let abs_client = AbstractClient::builder(mock.clone()).build()?;
// Add initial balance
mock.add_balance(&sender, coins(1_000_000, "uaxone"))?;
// Create publisher account
let publisher = abs_client
.account_builder()
.namespace(namespace)
.build()?
.publisher()?;
// Publish and install the app
publisher.publish_app::<MyContractInterface<_>>()?;
let app = publisher
.account()
.install_app::<MyContractInterface<_>>(
&MyContractInstantiateMsg::default(),
&[]
)?;
Ok(TestEnv {
abs: abs_client,
app,
})
}
}
Test Patterns
Success Case Testing
rust
#[test]
fn successful_install() -> anyhow::Result<()> {
let env = TestEnv::setup()?;
let app = env.app;
let config = app.config()?;
assert_eq!(config, ConfigResponse { /* expected */ });
Ok(())
}
#[test]
fn successful_execute() -> anyhow::Result<()> {
let env = TestEnv::setup()?;
let app = env.app;
// Execute action
app.do_something("param")?;
// Verify state changed
let result = app.query_something()?;
assert_eq!(result.value, expected_value);
Ok(())
}
Error Case Testing
rust
#[test]
fn unauthorized_access() -> anyhow::Result<()> {
let env = TestEnv::setup()?;
let app = env.app;
// Call as non-admin
let err: MyContractError = app
.call_as(&Addr::unchecked("not_admin"))
.admin_only_action()
.unwrap_err()
.downcast()
.unwrap();
assert_eq!(err, MyContractError::Unauthorized {});
Ok(())
}
Balance and Fund Testing
rust
#[test]
fn balance_operations() -> anyhow::Result<()> {
let env = TestEnv::setup()?;
let account = env.app.account();
// Add balance to account
let funds = coins(100, "uaxone");
account.add_balance(&funds)?;
let balances = account.query_balances()?;
assert_eq!(balances, funds);
// Add balance to any address
let mock_env = env.abs.environment();
mock_env.add_balance(&env.app.address()?, funds.clone())?;
Ok(())
}
Multiple User Testing
rust
#[test]
fn multi_user_scenario() -> anyhow::Result<()> {
let env = TestEnv::setup()?;
let mock = env.abs.environment();
// Create additional users
let alice = mock.addr_make("alice");
let bob = mock.addr_make("bob");
mock.add_balance(&alice, coins(1000, "uaxone"))?;
mock.add_balance(&bob, coins(1000, "uaxone"))?;
// Test as different users
env.app.call_as(&alice).user_action()?;
env.app.call_as(&bob).user_action()?;
Ok(())
}
Running Tests
bash
# Run all unit tests cargo make test-unit # Run all tests (includes integration) cargo make test # Run tests with coverage cargo make test-coverage # Run specific test cargo test test_name -- --nocapture
Coverage Analysis
After running cargo make test-coverage, coverage is saved to lcov.info.
Viewing Coverage
bash
# Generate HTML report (requires lcov) genhtml lcov.info --output-directory coverage/ # Or use cargo-llvm-cov directly cargo llvm-cov --html --open
Coverage Goals
| Component | Target |
|---|---|
| Handlers | 100% |
| State operations | 100% |
| Error paths | 90%+ |
| Message parsing | 80%+ |
| Overall | 90%+ |
Test Checklist
For Each Handler
- • Happy path test
- • All parameter variations tested
- • Error cases tested (unauthorized, invalid input, etc.)
- • State changes verified
- • Events/attributes checked if applicable
For Queries
- • Empty state returns appropriate response
- • Populated state returns correct data
- • Pagination works correctly (if applicable)
- • Non-existent items handled gracefully
For State
- • Initial state after instantiation
- • State updates correctly after execute
- • State persists across operations
Best Practices
- •Isolate tests - Each test should set up its own environment
- •Test business logic - Focus on contract behavior, not framework
- •Use descriptive names -
test_increment_when_paused_fails - •Assert specific errors - Check exact error type, not just failure
- •Clean up - Don't leave state that could affect other tests
- •Test edge cases - Zero values, max values, empty strings