[01] RECOGNITION

Agents ship code.
QA.army bug bashes it.

Point QA.army at your web or mobile app and name the journey that cannot break. It runs the flow and returns a recording with reproduction steps when something fails.

Agent setup instruction
Use QA.army to establish regression coverage for this product.
Paste this instruction into your coding agent. Authentication and capability availability remain explicit.
Run your first test →

5 test runs free · 7-day trial · No credit card · MCP · CLI · A2A

[·] BUILT FOR TEAMS + AGENTS THAT SHIP DAILY
[02] CAPABILITIES

Test the journeys
scripts miss.

QA.army is built for critical flows that cross applications, accounts, and environments. Saved tests protect the journeys your team already knows; change-focused testing investigates what the latest pull request put at risk.

Halftone illustration — a courier bursting outward through radiating rays.
#1 · START

Start from any
coding agent.

You should not have to rebuild your test setup by hand every time an agent ships code. Give the agent one instruction; QA.army takes the handoff from there and keeps Workspace access explicit.

PROMPT
Use QA.army to establish regression coverage for this product.

# Your agent asks for sign-in only when it needs it.
Halftone illustration — a hooded classical figure with veiled composure.
#2 · CROSS BOUNDARIES

Test the whole
customer journey.

The valuable failures rarely stay inside one page. A GitHub App journey can require a commit, webhook delivery, and the resulting pull-request experience. QA.army is designed to follow that complete path.

JOURNEY GitHub App change
01 push a commit
02 observe webhook delivery
03 verify the pull-request experience
# One customer outcome, across application boundaries.
Halftone illustration — an ancient classical bust in profile, rendered as scanlines.
#3 · CORE COVERAGE

Protect what your
team already knows.

Core test groups are saved, durable journeys for the flows that must keep working: sign-in, checkout, onboarding, subscriptions, or the business-critical path unique to your product.

TEST GROUP Checkout
✓ signed-in purchase
✓ guest purchase
✓ failed-payment recovery
# Durable coverage for journeys worth protecting.
Halftone illustration — an antique pocket watch haloed by a bursting sun.
#4 · CHANGE-FOCUSED

Investigate what
the pull request changed.

Dynamic pull-request tests are the complementary model: read the description and diff, plan fresh checks around the change, and target gaps outside saved coverage. This capability is not generally available yet.

CHANGE authentication callback updated
FOCUS OAuth return path
FOCUS session recovery
FOCUS existing sign-in regressions
# Fresh investigation, grounded in the change.
Halftone illustration — a peacock with a fully fanned kaleidoscopic tail.
#5 · EVIDENCE

Get a bug report,
not a red dot.

A useful failure names the broken expectation and brings the proof: screenshots, a recording, reproduction steps, and the console or network evidence that explains what happened.

FAILED Intro offer after 2FA
ATTACHED screenshot + recording
REPRO 4 deterministic steps
EVIDENCE console + network context
# Clear enough to act on without rerunning by hand.
Halftone illustration — a celestial diagram with a radiant sun over stone steps.
#6 · ISOLATION

One attempt.
One isolated session.

Every run attempt gets its own execution session. Access stays bound to the Workspace, project, test, and run that QA.army resolved—never to scope chosen by the testing agent.

BOUND workspace → project → test → run
✓ isolated execution session
✓ run-scoped tools and evidence
✓ clean cancellation and termination
# The agent cannot switch itself into another scope.
[07] VOICES
“Saved tests protect what your team already knew to write down. Change-focused tests investigate what the pull request put at risk.”
— the QA.army coverage model
[08] FAQ

FAQ

What is QA.army, in one sentence?

QA.army is the agent-native QA product for critical journeys across web, mobile, and desktop. Saved Web-test authoring is available now; scheduled monitoring and native mobile or desktop execution are not generally available yet.

How do I start from my coding agent?

Paste the QA.army setup instruction above. Your agent can begin the handoff, define seven critical journeys, and request sign-in or an authorized Workspace selection when required.

What are core test groups and dynamic tests?

Core test groups are saved, durable journeys. Dynamic tests plan fresh coverage around a pull-request description and diff. Dynamic pull-request testing is not generally available yet.

What belongs in a useful result?

PASSED means the required expectations held. FAILED means the product broke an expectation. ERROR means QA.army could not reliably determine the product result. A useful failure includes screenshots, a recording, reproduction steps, and relevant console or network evidence.

Does it test mobile and desktop?

Web, mobile, and desktop are all part of QA.army. Native mobile and desktop execution are not generally available yet.

How are teams and data separated?

Workspace is the boundary for access and billing. Runs, tests, evidence, and agent access remain scoped to the authorized Workspace and project.

# product

Laura

Product Manager

The intro-offer bottom sheet stopped showing after 2FA login in dev.

Nat

CEO

@qa.army can you check?

QA.army app icon
QA.armyAPP

Checked the same journey across all three environments.

Production passes
Staging passes
Dev fails on 9.5.8-release

API responses and feature flags match. The frontend stops before mounting IntroOfferSheet. Release notes show the auth callback changed in this build.

Pick a priority, then log the ticket.

When a regression is caught, QA.army returns the result, screenshots, and the evidence needed to reproduce it.
[09] INVITATION

One instruction begins
the QA handoff.

Provide the QA.army setup instruction to your coding agent. It will define seven critical journeys and confirm what is currently ready to run and monitor.

We use your email to begin the authorized product sign-in.