using-toolkit

Routes incoming tasks to the appropriate skills and sizes the process to the change.

Updated Aug 3, 2026
One-click install
npx skills add https://github.com/m-de-graaff/skills --skill using-toolkit-m-de-graaff
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: using-toolkit
Source: https://github.com/m-de-graaff/skills/tree/main/skills/using-toolkit
Command: npx skills add https://github.com/m-de-graaff/skills --skill using-toolkit-m-de-graaff

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? When a task arrives, it is unclear which skills should apply and in what order, so work either gets buried in unnecessary process or skips the design, planning, and review steps it actually needs. This Skill acts as a router that matches the observable characteristics of the work to the right chain of skills. ## Core Features & Use Cases - Task routing: Maps what is observably true about the work (bug, performance change, multi-file refactor, code removal) to an ordered chain of skills such as systematic-debugging, writing-plans, or requesting-code-review. - Proportional process sizing: Scales the workflow to the blast radius of the change, from a one-line fix that needs no plan to an auth refactor that needs design, worktrees, and two-way review. - Domain reference loading: Directs on-demand loading of domain skills (accessibility, security-patterns, database-migrations, and more) at the moment the work reaches that area. - Use Case: A user reports that a dashboard feels slow. The router matches this to the performance row and directs the chain measuring-performance, then optimization-loop, then verification-before-completion, then git-workflow. ## Quick Start Ask the assistant to route a new task, for example: "Route this request and tell me which skills apply: fix the flaky login test and then refactor the session handling."

Frequently Asked Questions about using-toolkit

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

FAQPage Schema
How do I decide which skills to use for a coding task?▼

Match what is observably true about the work against the routing table rather than the wording of the request. A bug routes to systematic-debugging first, multi-file ordered work routes to writing-plans then executing-plans, and a one-file reversible change needs no chain at all.

What is skill routing in an AI coding workflow?▼

Skill routing is the step of naming the chain of skills that apply to a task and invoking each one before doing the work it governs. Reading the router alone is not routing; the output of routing is an actual invocation of each selected skill.

When should a task skip planning and design steps?▼

Skip the heavier tiers when the change is one file, reversible, and the user already stated exactly what they want, or when the request is a question about existing code with no change. When two tiers look plausible, take the lower one and say what was skipped.

Does the router handle tasks that grow or change scope mid-work?▼

Yes. When work gets bigger than it looked, stop and re-size the tier explicitly. When work moves into a new domain such as auth or UI, re-check the domain reference table even if the tier stays the same, since new rows may now apply.

When should domain skills like accessibility or security be loaded?▼

Load a domain reference at the moment the work first reaches that area, not only at the start. Any user-facing UI change loads accessibility, and any work touching auth, user input, uploads, secrets, or payments loads security-patterns, regardless of task size.