VividKit Crystal clear AI coding
VI
Choose Your Setup Get Started with AgentKit Continue with ClaudeKit Migrate from ClaudeKit to AgentKit Clean cutover from ClaudeKit What Is AgentKit? How AgentKit Works AgentKit Kits, Skills, Agents & Hooks AgentKit Runtime Support AgentKit Desktop App AgentKit Configuration AgentKit CLI Quick Reference AgentKit Skills Cheatsheet AgentKit Workflows AgentKit Statusline Update AgentKit CLI & Kits AgentKit Helper Clean reinstall of AgentKit AgentKit Troubleshooting What is ClaudeKit? ClaudeKit CLI Guide Command Finder Commands Cheat Sheet Workflow Recipes CLI Quick Reference Claude Mechanics Permission Modes ClaudeKit Hooks Coexistence ClaudeKit x Codex CK → Other coding tools How ClaudeKit Works Inside ClaudeKit Finding Your Unknowns Session Recovery Fix from Logs UI/UX Guide Accounts Switcher IDE Config Remote Control Codex App System Diagramming UI Review Gate Promotions Support VividKit

Quick Ref / Command

Engineer Kit

/ak:coding-level

Set coding experience level

Persist a 0–5 codingLevel preference so future supported sessions tailor explanation depth, code complexity, and response format.

01

Invoke explicitly

02

Persist config

03

Start supported session

04

Inject guidance

Rule 01

Match explanation depth to the reader, not the author.

Rule 02

Level 5 is the documented maximum-efficiency default.

Rule 03

The setting persists in .claude/.ck.json and is injected by supported session-start hooks.

Rule 04

Coding level changes communication style; it does not change repository rules, safety boundaries, or authorization.

Execution Map

Core lanes from input to output, aligned with the command's real execution path.

01

Start

Input, scope, route

  1. 1 Invoke explicitly Run /ak:coding-level with one valid integer from 0 through 5; the Skill disables automatic model invocation.
  2. 2 Persist config Update codingLevel in .claude/.ck.json while preserving unrelated project configuration.
02

Work

Agent / skill execution

  1. 3 Start supported session On session start, the Engineer Hook reads the saved value and defaults to level 5 when codingLevel is absent.
  2. 4 Inject guidance The Hook maps the selected level to the matching communication rules and style name.
03

Verify

Gate, review, validation

  1. 5 Shape responses Lower levels add analogies, why-before-how, common mistakes, and takeaways; higher levels emphasize concise trade-offs, architecture, risk, and strategy.
  2. 6 Optional output style Claude Code can also select coding-level-0-eli5 through coding-level-5-god; same-named user styles are preserved with a warning during routine updates.

Syntax, arguments, subcommands

How to invoke the skill: syntax, positional arguments, shared options, then each subcommand with its own syntax and outcome.

Syntax

/ak:coding-level [0-5]

Arguments

[0-5] Required

Level

One integer from 0 through 5. The skill stores it in .claude/.ck.json and does not change repo rules or safety.

/ak:coding-level 2

Modes / Routes

0

ELI5

Research:
Analogies, no jargon
Review:
Step-by-step guardrails
Validate:
Key takeaways
1

Junior

Research:
Concepts explained
Review:
WHY before HOW
Validate:
Common mistakes
2

Mid-Level

Research:
Design patterns
Review:
System thinking
Validate:
Trade-offs
3

Senior

Research:
Architecture context
Review:
Business trade-offs
Validate:
Concise rationale
4

Tech Lead

Research:
Risk assessment
Review:
Business impact
Validate:
Strategy
5

God Mode

Research:
Maximum efficiency
Review:
Minimal exposition
Validate:
Direct output

Sample Prompt

Concrete invocations for each flag, subcommand, mode, or route available in the live workflow.

Junior teaching mode Recommended 
/ak:coding-level 1
Use when:
Explanations feel too terse and you want WHY, common mistakes, and takeaways in later sessions.
Expected:
The Skill stores codingLevel: 1 in .claude/.ck.json; after a supported fresh session starts, responses should explain concepts clearly and teach why before how.
Mid-level system thinking 
/ak:coding-level 2
Use when:
You want less beginner scaffolding and more design patterns, system thinking, and explicit trade-offs.
Expected:
The persisted level becomes 2, so supported sessions should frame answers around patterns, testability, system effects, and practical trade-offs.
Tech-lead framing 
/ak:coding-level 4
Use when:
You need risk assessment, business impact, strategy, dependencies, and stakeholder decisions to be foregrounded.
Expected:
The config records codingLevel: 4; future supported responses should compress basics and emphasize risks, strategy, business impact, and coordination decisions.
Maximum efficiency 
/ak:coding-level 5
Use when:
You want the documented default expert profile with minimal exposition and direct output.
Expected:
codingLevel is persisted as 5, so supported fresh sessions should use maximum-efficiency expert output unless another instruction asks for more explanation.

Handled Scope

  • Invoke explicitly
  • Persist config
  • Start supported session
  • Inject guidance
  • Shape responses
  • Optional output style

Next