archflow-agent-qa-engineer

Tests web, mobile, and backend code against the project's declared stack and design system.

28|4|Updated Mar 12, 2026
One-click install
npx skills add https://github.com/AZidan/archflow --skill archflow-agent-qa-engineer-azidan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: archflow-agent-qa-engineer
Source: https://github.com/AZidan/archflow/tree/main/adapters/generic/.agents/skills/archflow-agent-qa-engineer
Command: npx skills add https://github.com/AZidan/archflow --skill archflow-agent-qa-engineer-azidan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code shipped by AI agents often passes a quick glance but breaks under real usage, drifts from the API contract, or violates the project's design system. This Skill runs a structured QA gate after implementation, catching defects, contract drift, and design anti-patterns before a story reaches acceptance review. ## Core Features & Use Cases - Stack-aware test authoring: Reads the declared stack from .archflow/project-settings.yaml or detects test tooling from repo manifests, then writes unit, integration, and E2E tests in the project's own runners. - Design system enforcement: Runs the design system's anti-pattern checklist as a blocking review gate, reporting violations with file, line, and the correct vocabulary term. - API contract verification: Validates every touched endpoint against the API contract field-for-field, treating drift as a failure. - Use Case: After ui-engineer and api-engineer finish a story, this agent writes and runs the test suite, saves findings to docs/qa-reports/, records issues in the release file, and reports pass/fail to the orchestrator. ## Quick Start Ask the agent to run the QA gate for the current story, writing tests in the project's declared stack and reporting results to docs/qa-reports.

Frequently Asked Questions about archflow-agent-qa-engineer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run QA testing on a story after implementation?

Invoke this agent after ui-engineer and api-engineer complete a story. It reads the declared stack, writes tests in the project's own runners, executes them, and saves a report to docs/qa-reports before the story proceeds to pm-reviewer.

What test frameworks does the QA engineer agent support?

It carries no framework of its own and uses whatever the project declares in project-settings.yaml. When the stack is unset, it detects tooling from package.json, pyproject.toml, go.mod, build.gradle, and existing test directories, then asks for confirmation.

Can the QA agent install a missing test framework?

No. Installing a test framework, browser, or emulator is treated as a project-shaping decision reserved for the user. The agent names what it would install and why, then waits for approval.

Why does a story fail review even when all tests pass?

The agent runs two gates: the test suite and the design system anti-pattern checklist. Any design-system violation or API contract drift fails the review regardless of green tests, and the story returns to the implementation agent.

Does the QA engineer fix the bugs it finds?

No. It records every finding as a structured issue in the release file with severity and location, then stops. Fixes go back to ui-engineer or api-engineer so the checking agent never clears its own findings.