SKILL.md
Test Runner
Execute and manage Rust tests for the self-learning memory project.
Commands
# Unit tests (fast, <30s)
cargo nextest run --lib
# Integration tests
cargo nextest run --test '*'
# Full suite
cargo nextest run --all
cargo test --doc # doctests (nextest unsupported)
# Quality gates (coverage threshold 90%)
./scripts/quality-gates.sh
Test Categories
| Category |
Command |
Scope |
| Unit |
cargo nextest run --lib |
Individual functions |
| Integration |
cargo nextest run --test '*' |
End-to-end workflows |
| Doc |
cargo test --doc |
Documentation examples |
| Mutation |
cargo mutants -p do-memory-core |
Test effectiveness |
Best Practices
- Isolation: Each test independent
- AAA pattern: Arrange-Act-Assert
- Single responsibility: One behavior per test
- Naming:
test<function><scenario>_<expected>
- Speed: <1s per unit test
Async Testing
#[tokio::test]
async fn test_episode() { let result = async_fn().await; }
#[tokio::test(start_paused = true)]
async fn test_timeout() { /* paused clock */ }
#[tokio::test(flavor = "multi_thread", worker_threads = 4)]
async fn test_concurrent() { /* parallel ops */ }
| Bad |
Good |
std::thread::sleep() |
tokio::time::sleep().await |
Missing .await |
Always await async |
Troubleshooting
| Issue |
Fix |
| Race conditions |
cargo test -- --test-threads=1 |
| redb lock errors |
Separate DB per test |
| DB connection refused |
Check TURSOURL/TURSOTOKEN |
Coverage
cargo llvm-cov --html --output-dir coverage
References
- ADR-033: Modern Testing Strategy
Related skills: test-patterns for test design, test-optimization for performance-focused test work, and test-fix for failing-suite diagnosis.
