Paths: File paths (shared/ , references/ , ../ln-* ) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root. If shared/ is missing, fetch files via WebFetch from https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path} .

MANDATORY READ: Load shared/references/ci_tool_detection.md — test framework registry, compact flags (--tb=short -q , --reporter=json ), and normalization chain.

ln-782-test-runner

Type: L3 Worker Category: 7XX Project Bootstrap

Purpose

Detects test frameworks, executes all test suites, and reports results including pass/fail counts and optional coverage.

Scope:

• Auto-detect test frameworks from project configuration

• Execute test suites for all detected frameworks

• Parse test output for pass/fail counts

• Generate coverage reports when enabled

Out of Scope:

• Building projects (handled by ln-781)

• Container operations (handled by ln-783)

• Writing or fixing tests

When to Use

Scenario Use This Skill

Standalone-capable Yes

Standalone test execution Yes

CI/CD pipeline test step Yes

Build verification needed No, use ln-781

Workflow

Step 1: Detect Test Frameworks

Identify test frameworks from project configuration files.

Marker Test Framework Project Type

vitest.config.* Vitest Node.js

jest.config.* Jest Node.js

*.test.ts in package.json Vitest/Jest Node.js

xunit / nunit in *.csproj xUnit/NUnit .NET

pytest.ini / conftest.py pytest Python

*_test.go files go test Go

tests/ with Cargo.toml cargo test Rust

Step 2: Execute Test Suites

Run tests for each detected framework.

Framework Execution Strategy

Vitest Run in single-run mode with JSON reporter

Jest Run with JSON output

xUnit/NUnit Run with logger for structured output

pytest Run with JSON plugin or verbose output

go test Run with JSON output flag

cargo test Run with standard output parsing

Step 3: Parse Results

Extract test results from framework output.

Metric Description

total Total number of tests discovered

passed Tests that completed successfully

failed Tests that failed assertions

skipped Tests marked as skip/ignore

duration Total execution time

Step 4: Generate Coverage (Optional)

When coverage enabled, collect coverage metrics.

Framework Coverage Tool

Vitest/Jest c8 / istanbul

.NET coverlet

pytest pytest-cov

Go go test -cover

Rust cargo-tarpaulin

Coverage Metrics:

Metric Description

linesCovered Lines executed during tests

linesTotal Total lines in codebase

percentage Coverage percentage

Step 5: Report Results

Return structured results to orchestrator.

Result Structure:

Field Description

suiteName Test suite identifier

framework Detected test framework

status passed / failed / error

total Total test count

passed Passed test count

failed Failed test count

skipped Skipped test count

duration Execution time in seconds

failures Array of failure details (test name, message)

coverage Coverage metrics (if enabled)

Error Handling

Error Type Action

No tests found Report warning, status = passed (0 tests)

Test timeout Report timeout, include partial results

Framework error Log error, report as error status

Missing dependencies Report missing test dependencies

Options

Option Default Description

skipTests false Skip execution if no tests found

allowFailures false Report success even if tests fail

coverage false Generate coverage report

timeout 300 Max execution time in seconds

parallel true Run test suites in parallel when possible

Critical Rules

• Run all detected test suites - do not skip suites silently

• Parse actual results - do not rely only on exit code

• Include failure details - provide actionable information for debugging

• Respect timeout - prevent hanging on infinite loops

Monitor Integration (Claude Code 2.1.98+)

MANDATORY READ: Load shared/references/monitor_integration_pattern.md

For test suites expected to run >30 seconds, use Monitor to stream failures in real-time:

Pattern Command timeout_ms

Stream failures {test_command} 2>&1 | grep --line-buffered -E 'FAIL|Error|✗|AssertionError'

timeout * 1000 (from Options)

Full output {test_command} 2>&1

timeout * 1000

React to first failure immediately while remaining tests continue running.

Fallback: if Monitor is unavailable (Bedrock/Vertex), use Bash(run_in_background=true) .

Definition of Done

• All test frameworks detected

• All test suites executed

• Results parsed and structured

• Coverage collected (if enabled)

• Results returned to orchestrator

Reference Files

• Parent: ../ln-780-bootstrap-verifier/SKILL.md

Version: 2.0.0 Last Updated: 2026-01-10