volter-ai/Runhuman QA Test

Run human-powered QA tests in your CI/CD pipeline. Get real human feedback on UI, UX, and user flows.

Trust Signals

name	description	required	default
api-key	Runhuman API key (starts with qa_live_). Store as secret!	yes	—
url	URL to test (must be publicly accessible)	yes	—
description	Test instructions for human tester	yes	—
output-schema	Optional JSON schema defining expected output structure. If omitted, only success/explanation are extracted.	no	—
api-url	Runhuman API base URL	no	https://runhuman.com
target-duration-minutes	Target test duration in minutes (1-60, default: 5)	no	5
allow-duration-extension	Allow tester to request more time (true/false)	no	false
max-extension-minutes	Maximum extension minutes allowed (number or false)	no	—
additional-validation-instructions	Custom instructions for the AI when validating test results (e.g., "Ignore minor CSS issues")	no	—
can-create-github-issues	Auto-create GitHub issues from tester findings (true/false)	no	false
screen-size	Screen size for testing. Use preset (desktop, laptop, tablet, mobile) or custom JSON {"width": 1920, "height": 1080}	no	desktop
fail-on-error	Fail workflow if test fails (default: true)	no	true
fail-on-timeout	Fail workflow if test times out before completion (default: false)	no	false

name	description
status	Job status (completed, error, incomplete, abandoned)
success	Boolean - true if test succeeded, false otherwise
result	Full result object as JSON string
explanation	Tester explanation/findings
data	Extracted data as JSON string
cost-usd	Test cost in USD (e.g., 0.396)
duration-seconds	Test duration in seconds
job-id	Job ID for reference
tester-alias	Anonymous tester name (e.g., Phoenix)
tester-avatar-url	Tester avatar URL
tester-color	Tester color for UI theming
tester-data	Playwright session data as JSON (screenshots, video, console logs, network requests)
tester-response	Raw tester response text
timed-out	Boolean - true if the test timed out before completion