archflow-agent-pm-reviewer

Validates story acceptance criteria by running end-to-end tests with existing project tooling.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code that passes QA can still fail to deliver what a user story promised. This Skill acts as the final product acceptance gate, verifying each acceptance criterion against the running software as a user would, and producing an actionable ACCEPTED, REJECTED, PARTIALLY ACCEPTED, or BLOCKED verdict. ## Core Features & Use Cases - Criterion-to-test mapping: Reads acceptance criteria from the active release file and writes one named end-to-end test per criterion using the tooling the project already has (Playwright, Cypress, Detox, XCUITest, Espresso, and similar). - Honest verdicts with evidence: Distinguishes code defects from environment problems and broken tests, captures screenshots or API responses, and writes a structured report to docs/acceptance-reports/. - Persistent issue tracking: Records every finding as a structured issue in the release YAML file so defects survive context compaction and reach the implementation agent intact. - Use Case: After qa-engineer passes a login story, this Skill runs the project's Playwright suite against each acceptance criterion, finds the error toast renders instead of the promised modal, and files a blocking P1 issue with reproduction steps. ## Quick Start Ask the agent to run the pm-reviewer acceptance review for story US-12 in the active release and report the verdict.

Frequently Asked Questions about archflow-agent-pm-reviewer

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

FAQPage Schema
How do I verify acceptance criteria with end-to-end tests?

Map each acceptance criterion to one named test that drives the app as a user would, asserting on the specific behavior the criterion names. Run the suite with the project's own test command and capture screenshots or API responses as evidence for each result.

What end-to-end testing tool should I use for my project type?

Match the tool to the project type: Playwright or Cypress for web, Detox or Maestro for React Native, XCUITest for native iOS, Espresso for native Android, and the project's HTTP test stack for backend-only. Always prefer tooling already installed over introducing a new framework.

What is the difference between QA testing and product acceptance review?

QA engineering proves the code works technically, while acceptance review verifies the software does what the story promised from a user's perspective. The acceptance reviewer runs after QA passes and judges each criterion against the running product, not the implementation.

What happens when acceptance tests cannot run at all?

The reviewer returns BLOCKED rather than REJECTED or ACCEPTED, listing which tools would suit the project type with install commands so the user can choose. The story must not be merged until acceptance testing actually executes.

Can the reviewer install Playwright or other test frameworks automatically?

No. Installing any test framework, browser, simulator, or emulator requires explicit user approval first, since choosing test tooling is a project-shaping decision. Without approval and no existing tooling, the outcome is BLOCKED.