How to use the Rust Testing Patterns agent skill (rust-testing)

What the rust-testing skill makes the agent do

An agent skill is a folder with a SKILL.md file that gives an agent extra instructions for a specific kind of task. When this skill is active, the agent approaches Rust testing in a structured way: it identifies the target code, writes a test first, mocks dependencies where needed, and works through a red-green-refactor cycle.

The skill asks the agent to run the tests and confirm they fail before writing the implementation (RED), write minimal code to make them pass (GREEN), then improve the code while keeping tests green (REFACTOR). At the end it suggests checking coverage with cargo-llvm-cov, targeting 80% or more.

Install the skill from the ZIP

Download the ZIP for the skill from this page and unzip it. The skill is a single file, SKILL.md. Put the unzipped skill folder in one of these locations:

• Personal skills: ~/.claude/skills/
• Project skills: .claude/skills/ inside your Rust project
• Claude apps: upload the skill under Settings → Capabilities → Skills

Check what you need before you start

The skill assumes a Rust toolchain with cargo is installed. The examples reference the crates rstest, proptest, mockall and tokio, and cargo-llvm-cov is optional and only needed for coverage checks. The skill does not run scripts and does not use the network.

• Rust toolchain with cargo
• rstest, proptest, mockall, tokio (referenced in examples)
• cargo-llvm-cov (optional, for coverage)

Ask for Rust tests in natural language

You do not need a special command. Describe the Rust work you want and mention tests, TDD or coverage. The skill is most useful when the request matches its scope: writing new functions, adding coverage to existing code, property tests for input validation, or following a TDD workflow.

• Write a failing test for a new add function in src/math.rs, then implement it using TDD.
• Add unit tests for User::new in src/user.rs, including the invalid email case.
• Add property-based tests with proptest for my encode and decode functions.
• Mock the UserRepository trait with mockall and test UserService::get_user.
• Parameterize this string length test with rstest for empty, short and long inputs.
• Check test coverage for this crate with cargo-llvm-cov and aim for 80%.

Patterns the agent will follow

The skill includes reference examples for each testing style. The agent uses them as a consistent pattern rather than inventing a new layout each time.

• Unit tests in a #[cfg(test)] module with #[test], and assertion macros such as assert_eq! and assert!.
• Error and panic tests: checking Result errors with matches!, returning Result from tests, and #[should_panic] with an expected message.
• Integration tests in a tests/ folder, where each file is a separate test binary, with shared helpers in tests/common/.
• Async tests using #[tokio::test] and tokio::time::timeout.
• rstest for #[case] parameterized tests and #[fixture] fixtures.
• proptest for property tests and custom strategies, and mockall's #[automock] for trait mocks.

Tips for getting good results

The skill gives the agent a pattern, but you still review the tests. Ask the agent to show the failing test output before implementation, since the skill's RED step depends on it. If you only need a syntax reminder, the file is mostly examples, so reading the reference directly may be faster.

• Name the file and function you want tested so the agent knows the target code.
• Ask for the RED step explicitly if you want to see the failing test first.
• Request cargo-llvm-cov only if it is installed, since it is optional.
• Review mocks and property tests to confirm they check behaviour you care about.

Safety notes

The skill runs no scripts and makes no network calls, according to its own listing. Tests you ask the agent to run still execute code in your project through cargo, so read the test and code changes before running them on sensitive data or production systems.