using-workflow

Routes development requests through a deterministic skill-selection workflow from design to branch integration.

Updated Aug 3, 2026
One-click install
npx skills add https://github.com/Yvesdefaria/GymLab --skill using-workflow-yvesdefaria
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: using-workflow
Source: https://github.com/Yvesdefaria/GymLab/tree/main/gymlab-app/.agents/skills/using-workflow
Command: npx skills add https://github.com/Yvesdefaria/GymLab --skill using-workflow-yvesdefaria

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When building features or fixing bugs, it is easy to skip design, planning, testing, or review steps and jump straight into code. This Skill enforces a deterministic workflow that classifies each request and automatically routes it through the right process skills in the correct order, so nothing gets skipped. ## Core Features & Use Cases - Automatic skill routing: Classifies requests as spike, bounded, or architectural and chains the required skills (brainstorming, writing-plans, TDD, execution, review, finishing) with non-negotiable transition rules. - Hard approval gates: Blocks any code from being written before explicit design approval, and mandates writing-plans as the only next step after an approved spec. - Situational skill triggers: Maps concrete events (failing tests, UI work, SEO, accessibility, debugging) to the exact skill that must handle them. - Use Case: A user asks to add a new training-log feature to a React app. The Skill classifies it as architectural, requires a spec and plan first, then drives TDD-based execution, code review, and a user-decided merge or PR at the end. ## Quick Start Ask the assistant to start a new feature or project and have it route the entire process through the workflow, announcing each stage and required skill as it goes.

Frequently Asked Questions about using-workflow

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

FAQPage Schema
How do I enforce a design-before-code workflow for new features?

Classify the request as spike, bounded, or architectural, then require explicit design approval before any implementation. Architectural work must produce a written spec followed by a bite-sized plan with TDD steps before execution begins.

What is the correct order of steps for building a new feature?

The flow is brainstorming and classification, design approval, spec for architectural work, a written plan, then execution via subagent-driven development or executing-plans, and finally finishing-a-development-branch where the user chooses merge, PR, or keep.

When should test-driven development be applied in this workflow?

TDD applies to every feature, bugfix, or refactor before any production code is written. Each task in the plan follows the red-green-refactor cycle, and a test must be seen failing before it is considered valid.

What happens when a test fails or a bug appears mid-workflow?

Any bug, failing test, or unexpected behavior triggers systematic-debugging and stops forward progress. The process requires root-cause analysis before fixes, and after three failed fix attempts the architecture itself is questioned.

Can I skip the planning step for small changes?

Small bounded tasks still require a short design in chat and explicit approval, but they can go straight to TDD without a written plan document. Simple means short design, not zero design.