Run Tests

What it does

Spawns five independent test agents in parallel — unit, integration, end-to-end, regression, and load — against the implemented tasks. Each agent is independent of the developer who wrote the code and writes its own verdict report. Any RED verdict blocks the pipeline until a fix iteration is complete.

When to use it

Run after all tasks for a REQ have status: implemented and dev-reports are on disk. All five test tiers run in one command; the pipeline does not advance to reviews until every tier is GREEN.

Inputs

• ›REQ id
• ›TASK-N-dev-report.md files (status: implemented)

Outputs

• ›TASK-N-test-unit-report.md
• ›TASK-N-test-integration-report.md
• ›TASK-N-test-e2e-report.md
• ›TASK-N-test-regression-report.md
• ›TASK-N-test-load-report.md

Example

/eng-org:run-tests REQ-20260723-d904-01 — spawns five test agents in parallel; each writes a report with a GREEN or RED verdict; any RED triggers a fix iteration before reviews may start.