solo-qa

Verifies completed slices against done criteria and design before solo sign-off.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/scoots31/engineering-playbook --skill solo-qa-scoots31
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: solo-qa
Source: https://github.com/scoots31/engineering-playbook/tree/main/skills/solo-qa
Command: npx skills add https://github.com/scoots31/engineering-playbook --skill solo-qa-scoots31

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code-complete is not done. This Skill closes the gap between "the code compiles" and "the slice actually works as specified" by enforcing a two-part verification: active AI testing against the full slice spec, followed by mandatory human sign-off in the browser. ## Core Features & Use Cases - Active AI Verification: Walks each done criterion against the running slice in preview, cross-referencing rendered values against mock data files rather than trusting code inspection. - Design Fidelity Check: Compares the running implementation against the design file for layout, data rendering, edge states, and mock indicator badges. - Solo Sign-Off Gate: Presents a structured sign-off request the human must confirm in a live browser — it cannot be skipped, delegated, or performed by the AI. - Deliverable Acceptance: Runs a separate acceptance pass when all slices in a deliverable reach In Test, including UX impact checks for live-data integrations. - Use Case: After code review passes on slice SL-12, this Skill verifies the rendered value matches slot_target: 14.2 in the mock file, confirms design fidelity, then asks the solo to confirm all three done criteria in the browser before marking the slice In Test. ## Quick Start Run solo-qa on slice SL-12 to verify its done criteria against the running preview and request my sign-off.

Frequently Asked Questions about solo-qa

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

FAQPage Schema
How do I verify a slice is actually done and not just code-complete?

Run active verification against the running slice in preview: check each done criterion by observing rendered output, cross-reference values against mock data files, and confirm design fidelity. Then require human sign-off in a live browser before marking the slice In Test.

What is the difference between code review and QA verification?

Code review checks code quality and correctness by reading source. QA verification actively tests the running application — observing rendered values, edge cases, and design fidelity in the browser. This Skill only runs after code review has passed and logged its confirmation.

Can AI perform the final sign-off on a slice?

No. The solo sign-off is a required human gate that cannot be performed by the AI. The AI verifies criteria technically, but the human must confirm the slice looks and behaves correctly in a live browser before it moves to In Test.

What happens when QA finds a bug versus a missing requirement?

The Skill invokes a triage step that classifies the discovery as a bug, missing requirement, or regression, then routes it accordingly. A slice with an open missing requirement must never receive sign-off — the behavior must be defined or explicitly deferred first.

When does a slice move from In Test to Done?

A slice reaches In Test after solo sign-off, and reaches Done only when phase-level testing confirms the full phase works end-to-end. Both gates are required and neither can be skipped or substituted.