TestBox Runners — Comprehensive Reference

When to Use This Skill

• Running tests from the command line (CommandBox or BoxLang CLI)
• Configuring testbox in box.json
• Starting a file-watcher to auto-run tests on change
• Using --stream / StreamingRunner for SSE-based real-time output
• Invoking TestBox programmatically from CFML/BoxLang code
• Choosing the right runner for a use case (CLI, web, CI, programmatic)

---

Runner Overview

Runner	Use Case
CommandBox CLI	Standard development workflow; wraps testbox run
BoxLang CLI	BoxLang projects; binary at ./testbox/run
HTML Web Runner	Browser-based interactive test runner
Programmatic API	Embedding tests in application code or custom scripts
StreamingRunner	Real-time SSE-based output (TestBox 7+)

---

CommandBox CLI Runner

Basic Usage

# Run all tests configured in box.json
testbox run

# Override directory at runtime
testbox run runner="http://localhost:8080/tests/runner.cfm"

box.json Configuration

{
  "testbox": {
    "runner": "http://localhost:8080/tests/runner.cfm",
    "verbose": false,
    "labels": "",
    "excludes": "",
    "reporter": "min",
    "recurse": true,
    "bundles": "",
    "directory": {
      "mapping": "tests.specs",
      "recurse": true,
      "filter": ""
    },
    "options": {}
  }
}

Key CLI Flags

# Streaming real-time output (SSE)
testbox run --streaming

# Verbose output
testbox run verbose=true

# Specific reporter
testbox run reporter=json

# Run with labels
testbox run labels=integration

# Limit concurrent specs (parallel)
testbox run options.maxParallel=4

File Watcher

Start a persistent process that re-runs tests on file changes:

testbox watch

# Watch specific directory
testbox watch directory=models

# With streaming output
testbox watch --streaming

Watch is configured in box.json:

{
  "testbox": {
    "runner": "http://localhost:8080/tests/runner.cfm",
    "watchDelay": 500,
    "watchPaths": "**.cfc,templates/**"
  }
}

---

BoxLang CLI Runner

For BoxLang projects, TestBox ships a native binary runner at ./testbox/run.

Installation

TestBox CLI runner is automatically installed when TestBox is a dependency:

# Install testbox (adds ./testbox/run)
install testbox

All Flags

# Run all specs under tests/
./testbox/run

# Run specific directory
./testbox/run --directory=tests/unit

# Run specific bundles (comma-separated)
./testbox/run --bundles=tests.specs.MySpec,tests.specs.OtherSpec

# Streaming / real-time output
./testbox/run --stream

# Dry run — discover specs without executing
./testbox/run --dry-run

# Show only failed specs
./testbox/run --show-failed-only

# Show passed specs explicitly
./testbox/run --show-passed

# Show skipped specs explicitly
./testbox/run --show-skipped

# Show full stack traces
./testbox/run --stacktrace

# Set maximum failures before stopping
./testbox/run --max-failures=10

# Flag specs slower than N ms
./testbox/run --slow-threshold-ms=100

# Show top N slowest specs
./testbox/run --top-slowest=5

# Set reporter (min, json, junit, etc.)
./testbox/run --reporter=min

# Combine options
./testbox/run --directory=tests/unit --show-failed-only --slow-threshold-ms=50 --top-slowest=5

Exit Codes

Code	Meaning
0	All tests passed
1	One or more failures or errors

Use exit codes in CI/CD pipelines to gate deployments.

---

StreamingRunner (TestBox 7+)

The StreamingRunner emits specs in real time using SSE (Server-Sent Events), allowing the terminal or browser to show results as they complete instead of waiting for the full suite.

Via CommandBox

testbox run --streaming

Via BoxLang CLI

./testbox/run --stream

Programmatic Setup

var runner = new testbox.system.runners.StreamingRunner()
runner.run(
    directory: { mapping: "tests.specs", recurse: true },
    reporter:  "min",
    labels:    []
)

Web Endpoint (SSE)

The web-based streaming endpoint sends text/event-stream:

GET /testbox/runners/streamingrunner.bxm?reporter=min&directory=tests.specs

Browser-side:

const es = new EventSource( "/testbox/stream?reporter=min" )
es.onmessage = ( e ) => console.log( JSON.parse( e.data ) )

---

HTML Web Runner

The HTML runner is the classic browser-based test interface.

Quick Setup

http://localhost:8080/tests/runner.cfm

Create tests/runner.cfm (or tests/runner.cfm):

<cfscript>
    // Defaults: directory = /tests/specs
    new testbox.system.TestBox(
        directory: { mapping: "tests.specs", recurse: true }
    ).run()
</cfscript>

With Options

<cfscript>
    new testbox.system.TestBox(
        directory: { mapping: "tests.specs", recurse: true },
        reporter:  "simple",
        labels:    url.keyExists( "labels" ) ? url.labels : [],
        excludes:  [],
        options:   {}
    ).run()
</cfscript>

---

Programmatic Runner API

TestBox Construction

// By directory
var tb = new testbox.system.TestBox(
    directory: {
        mapping: "tests.specs",   // dot-notation mapping
        recurse:  true,
        filter:   ""              // optional glob filter
    }
)

// By specific bundles
var tb = new testbox.system.TestBox(
    bundles: [
        "tests.specs.unit.UserServiceSpec",
        "tests.specs.integration.APISpec"
    ]
)

// With reporter and labels
var tb = new testbox.system.TestBox(
    directory: { mapping: "tests.specs", recurse: true },
    reporter:  "json",
    labels:    [ "unit", "fast" ],
    excludes:  [ "slow" ]
)

Running

// Run and render output (returns and writes results to browser/stdout)
tb.run()

// Run and return raw result struct
var results = tb.runRaw()
// results.totalDuration, results.totalSpecs, results.totalPass,
// results.totalFail, results.totalError, results.totalSkipped

// Run via HTTP and return result string (for cross-server testing)
var output = tb.runRemote(
    testBundles: "tests.specs.MySpec",
    reporter:    "json",
    options:     {}
)
writeOutput( output )

Inspecting Raw Results

var results = new testbox.system.TestBox(
    directory: { mapping: "tests.specs", recurse: true }
).runRaw()

if ( results.totalFail > 0 || results.totalError > 0 ) {
    // CI: exit with failure
    systemOutput( "TESTS FAILED: #results.totalFail# failures, #results.totalError# errors" )
    abort
}

---

Auto-Load via Application.bx

In BoxLang projects, wire TestBox into the app lifecycle so you can hit /tests/ without a dedicated runner file:

// Application.bx
class {
    variables.this.name    = "MyApp"
    variables.this.mappings[ "/testbox" ] = expandPath( "/testbox" )
    variables.this.mappings[ "/tests"   ] = expandPath( "/tests"   )
}

---

Integration Testing with bx-web-support

Install the bx-web-support BoxLang module to run web requests from the CLI runner:

install bx-web-support
./testbox/run --directory=tests/integration

Integration tests can then make HTTP calls directly without a running web server.

---

CI/CD Integration

# GitHub Actions example
- name: Run TestBox Tests
  run: |
    ./testbox/run --directory=tests --reporter=junit --show-failed-only
  # Non-zero exit on failure automatically fails the step

# Jenkins / generic shell
./testbox/run --directory=tests --reporter=junit
if [ $? -ne 0 ]; then
  echo "Tests failed"
  exit 1
fi

---

Quick Reference

Task	Command
Run all tests	./testbox/run or testbox run
Stream output	./testbox/run --stream
Only failures	./testbox/run --show-failed-only
Dry run	./testbox/run --dry-run
Stack traces	./testbox/run --stacktrace
Slow spec report	./testbox/run --slow-threshold-ms=200 --top-slowest=10
Max failures	./testbox/run --max-failures=5
Specific directory	./testbox/run --directory=tests/unit
Specific bundles	./testbox/run --bundles=tests.specs.MySpec
JUnit for CI	./testbox/run --reporter=junit
Watch mode	testbox watch