Run Tests

run-tests

Execute tests based on project type. Uses centralized detection to determine the test framework and runs appropriate commands.

Usage

/run-tests

Runs all tests by default. You can also specify:

• A file path to run tests in a specific file
• A test name to run a specific test (if supported)

Instructions

Step 1: Discover Project Context

Use the Explore agent to discover project context:

Explore Prompt:

Discover project context for running tests. Find and read:

1. Root CLAUDE.md - Read CLAUDE.md at project root. All instructions are MANDATORY.
2. Relevant CLAUDE.md Files - Search **/CLAUDE.md for keywords: test, testing, jest, pytest, vitest, mocha, coverage
3. Project Type - Detect from package.json, pyproject.toml, go.mod, Cargo.toml, etc.

Return: Project type, package manager, test framework, test command, testing instructions

From the Explore results, extract:

• Project type and package manager
• Test framework
• Test command
• Any testing-specific instructions from CLAUDE.md

Step 2: Run Tests

Execute the test command from detection results:

Node.js:

# Using detected package manager
npm test
# or with specific file
npm test -- path/to/test.js

Python:

pytest
# or with specific file
pytest path/to/test.py
# or with specific test
pytest path/to/test.py::test_name

Go:

go test ./...
# or specific package
go test ./pkg/...
# with verbose
go test -v ./...

Rust:

cargo test
# specific test
cargo test test_name

Make:

make test

Java Maven:

mvn test
# or specific test
mvn test -Dtest=TestClass

Java Gradle:

./gradlew test
# or specific test
./gradlew test --tests TestClass

.NET:

dotnet test
# or specific test
dotnet test --filter "FullyQualifiedName~TestName"

Ruby:

bundle exec rspec
# or specific file
bundle exec rspec spec/models/user_spec.rb
# or specific line
bundle exec rspec spec/models/user_spec.rb:42

Step 3: Parse Results

After running tests, parse the output to identify:

• Total tests run
• Passed tests
• Failed tests
• Skipped tests
• Error messages for failures

Step 4: Report Results

Output in this format (see below).

Step 5 (Final): Evaluate Reusable Patterns

Run the pattern-evaluator agent to assess whether any reusable patterns (rules, skills, or agents) were discovered during this session and should be persisted.

Report Format

Output in this format:

## Test Results

**Framework**: [detected framework]
**Command**: [command executed]
**Status**: PASSED / FAILED

**Summary**:
- Total: X
- Passed: Y
- Failed: Z
- Skipped: W

[If failures, show details]

Detection Precedence

When multiple project files exist, use this order:

1. CLAUDE.md instructions - Explicit test command (highest priority)
2. package.json - scripts.test field (Node.js)
3. pyproject.toml - [tool.pytest] section (Python)
4. go.mod - Presence indicates Go project
5. Cargo.toml - Presence indicates Rust project
6. Gemfile - Presence indicates Ruby project
7. pom.xml / build.gradle - Java project
8. Makefile - test target (fallback)

If detection fails, ask the user:

Question: "Could not detect test framework. What command runs your tests?"
Options:
  - Provide test command
  - Cancel

Exit Code Handling

Interpret test runner exit codes:

Note: Don't rely solely on exit codes - always parse output for accurate counts, as some frameworks exit 0 with skipped tests.

Error Handling

Test Framework Not Installed

If the test command fails with "command not found":

Question: "Test framework not installed. What should I do?"
Options:
  - Show installation instructions
  - Try a different test command
  - Cancel

Installation commands by framework:

• Jest/Node: npm install --save-dev jest
• pytest: pip install pytest
• Go: Built-in, check go installation
• Rust: Built-in with cargo
• RSpec: bundle install or gem install rspec

No Tests Found

If tests run but find no test files:

No tests found. Checked:
- [list of paths searched based on framework conventions]

Question: "No tests found. What should I do?"
Options:
  - Specify test file/directory
  - Check different location
  - Cancel

Tests Timeout

If tests run longer than 5 minutes without completing:

Warning: Tests have been running for over 5 minutes.

Question: "Tests are taking a long time. How should I proceed?"
Options:
  - Continue waiting
  - Stop tests and show partial results
  - Run with --bail/fail-fast flag
  - Cancel

Missing Dependencies

If tests fail due to missing dependencies:

Tests failed due to missing dependencies.

Error: [error message]

Question: "Dependencies may not be installed. What should I do?"
Options:
  - Run install command (npm install, pip install, etc.)
  - Continue anyway
  - Cancel

Build/Compilation Failure

If tests fail before running due to build errors:

## Build Failed

Cannot run tests - compilation errors occurred.

[Build error output]

Fix the build errors first, then run `/run-tests` again.

Project Root Detection

Project root is determined by (in order):

1. Directory containing the detected project file (package.json, pyproject.toml, etc.)
2. Git repository root (git rev-parse --show-toplevel)
3. Current working directory

Examples

/run-tests
-> Runs all tests in project

/run-tests src/utils.test.ts
-> Runs tests in specific file

/run-tests --coverage
-> Runs tests with coverage (if supported)