Using Toolkit

<EXTREMELY-IMPORTANT> If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

This is not negotiable. This is not optional. You cannot rationalize your way out of this. </EXTREMELY-IMPORTANT>

Overview

Master catalog and dispatch system for all 64 toolkit skills. Ensures proper skill identification at conversation start.

How to Access Skills

In Claude Code: Use the Skill tool. When you invoke a skill, its content is loaded and presented to you — follow it directly. Never use the Read tool on skill files.

---

Phase 1: Skill Identification

Before ANY action: scan catalog, identify ALL matching skills (even 1% relevance), invoke via Skill tool.

---

Phase 2: Skill Prioritization

Invoke in order: (1) Process (brainstorming, planning) → (2) Specialist (frontend, backend, architect) → (3) Quality (TDD, review) → (4) Docs → (5) Design → (6) Ops → (7) Terminal (verification).

---

Phase 3: Skill Execution

Rigid skills: Follow exactly. Flexible skills: Adapt to context. Always invoke verification-before-completion before claiming done.

---

Available Skills (64 Total)

Core Skills (6)

Process & Workflow Skills (9)

Quality Assurance Skills (17)

Documentation Skills (5)

Design Skills (3)

Operations Skills (7)

Creative Skills (6)

Business Skills (3)

Document Processing Skills (3)

Frameworks & Languages Skills (3)

Productivity Skills (1)

Communication Skills (1)

---

Decision Table: Choosing the Right Skill

---

Workflow Patterns

---

Anti-Patterns

• Never skip skill check for "simple" tasks — always check catalog
• Never invoke skills from memory — use Skill tool for current version
• Never read skill files with Read tool — use Skill tool
• Never skip verification — always invoke verification-before-completion last
• Never respond before checking skills — skill check is ALWAYS first

---

Anti-Rationalization Guards

If you think "This is just simple / I need context first / Let me explore first / This doesn't need a skill / I remember this skill / The skill is overkill / I'll do one thing first" — STOP. Check for skills FIRST. No exceptions.

---

Integration Points

Key integrations: self-learning (project context), verification-before-completion (terminal checkpoint), circuit-breaker (stall safety), auto-improvement (effectiveness tracking), planning (most common first skill), resilient-execution (failure recovery).

---

Core Rules

Plan before coding. TDD always. Verify completion with evidence. Review before merge. Use subagents for parallel work. Self-learn continuously. Try 3 approaches before escalating. Report RALPH_STATUS in loops. Protect config files. Write specs with acceptance criteria.

---

Skill Types

Rigid (TDD, debugging, planning, verification, code-review, autonomous-loop, circuit-breaker, spec-writing, acceptance-testing): Follow exactly. Flexible (brainstorming, tech-docs, api-design, frontend, database, performance, security-review, prd-generation, laravel-specialist, php-specialist): Adapt to context.

---

Find Missing Skills

npx skills find [query] to search, npx skills add <owner/repo@skill> -g -y to install. Prefer 1K+ weekly installs from reputable sources.

---

Skill Type

RIGID — Skill checking is mandatory. Every task begins with a catalog scan. No exceptions. No rationalization.