cortex-test

name	cortex-test
description	Use when running automation tests, checking test results, or after code changes that need verification. Accepts optional domain parameter like cortex-test Data

Cortex Test

Unified dual-track test runner for Unreal C++ and Python MCP tests.

Usage

• cortex-test — run both tracks
• cortex-test Data — Unreal Cortex.Data+ tests only
• cortex-test python — Python MCP tests only
• cortex-test e2e — Python E2E tests (ensures editor running first)
• cortex-test benchmark — MCP benchmark tests (Layer 1 + Layer 2, requires editor)
• cortex-test scenarios — MCP scenario tests only (Layer 2)
• cortex-test stress — MCP stress tests only

Steps

1. Determine Test Track

Parse the argument:

• No argument → run both tracks
• Domain name (Data, Core, Graph, Blueprint, UMG, Level, Reflect, Editor, QA) → Unreal tests for that domain
• python → Python unit tests
• e2e → Python E2E tests
• benchmark → MCP benchmark (Layer 1 E2E + Layer 2 scenarios)
• scenarios → MCP scenario tests only
• stress → MCP stress tests only

2. Unreal C++ Tests

Run via PowerShell test runner:

powershell.exe -NoProfile -ExecutionPolicy Bypass -Command "& { Set-Location 'cli/Testing'; .\RunTests.ps1 'Cortex.<Domain>+' }"

For all domains, omit the filter or use Cortex+.

Quick status check (no re-run):

powershell.exe -NoProfile -ExecutionPolicy Bypass -Command "& { Set-Location 'cli/Testing'; .\RunTests.ps1 -ReportOnly }"

Zero warnings required — treat any warnings as failures.

3. Python MCP Tests

Check for test dependencies:

cd Plugins/UnrealCortex/MCP && uv run pytest tests/ -v

For E2E tests, first verify the editor is running (check Saved/CortexPort-*.txt exists). If not running, tell the user to start it or use cortex-editor.

3b. MCP Benchmark Tests

Requires running editor. First verify editor is running (same as E2E).

# benchmark — all TCP E2E + scenarios
cd Plugins/UnrealCortex/MCP && uv run pytest tests/test_e2e.py tests/test_level_e2e.py tests/test_editor_e2e.py tests/test_mcp_scenarios.py -v

# scenarios — cross-domain workflows only
cd Plugins/UnrealCortex/MCP && uv run pytest tests/test_mcp_scenarios.py -v -k "not stress"

# stress — performance tests only
cd Plugins/UnrealCortex/MCP && uv run pytest tests/test_mcp_scenarios.py -v -k stress

4. Report Results

Report:

• Which track(s) ran
• Pass/fail counts per track
• Any failures with file:line and error message
• Overall verdict: PASS or FAIL